Cu 31924004239087
Cu 31924004239087
Cu 31924004239087
UNIVERSITY LIBRARY
BOUGHT WITH THE INCOME OF THE SAGE ENDOWMENT FUND GIVEN IN 1891 BY
QC
An
451.W35
introduction to the study of spectrum
The
tine
original of
tliis
book
is in
restrictions in
AN INTRODUCTION
TO THE
X a
CO
O r > w m o
JO
c 3
D y
AN INTRODUCTION
TO THE
W.
MARSHALL WATTS
D.Sc. (LOND.), B.Sc. (Vict.), F.I.C.
IVITIf
CO.
PREFACE
Whatever
discoveries the century now opening may bring us, there can be no doubt of the enormous advance in scientific knowledge witnessed by the nineteenth century. That one hundred years ago
no traveOing by rail, no electric telegraph, no telephone, no electric or gas lighting, no photography, no chloroform not to mention such trifles as matches this shows that no mean advance has been made. But, amongst all the discoveries of the past century, from some points of view, at least there is none more important than the discovery of spectrum analysis, by means of which the composition of the sun and stars has been revealed and new elements have been discovered, and which is now the most powerful means
there was
we
possess
of
investigating
the
constitution
of matter.
Before
the discoveries of Bunsen and Kirchhoff no philosopher had ever ventured to think it possible that we should be able to analyze the
sun and
stars.
The
idea of
its
possibility
is
to these philosophers
at
constructed instrument, they recognized the presence of strontium amongst the burning material, and the thought occurred to them, " If it be possible to recognize a substance by the nature
any distance ? possible since i860 by "By the the introduction into the observatory of the spectroscope and the modern photographic plate, a great harvest has been gathered by many reapers. Spectroscopic astronomy has become a distinct and
of the light emitted a few miles
off,
also at
"
of
acknowledged branch of the science, possessing a large literature The more its own, and observatories specially devoted to it.
recent discovery of the gelatine dry plate has given a further great impetus to this modern side of astronomy, and has opened a path-
way into the unknown, of which even an enthusiast forty years ago would scarcely have dared to dream. " In no science, perhaps, does the sober statement of the results which have been achieved appeal so strongly to the imagination
vi
Preface.
and make so evident the almost boundless powers of the mind of man. By means of its light alone to analyze the chemical nature of a far distant body to be able to reason about its present state to measure within an English in relation to 'the past and future mile or less per second the otherwise invisible motion which it may have towards or from us; to do more to make even that which is darkness to our eyes light, and from vibrations which
;
in
our organs of sight are powerless to perceive, to evolve a revelation which we see mirrored some of the stages through which the
stars
may
of such achievements, however poor the form of words in which they may be described, is worthy to be regarded as the scientific
epic of the century."
""'
As
the
analysis
thus
opened up
importance.
will
and
But
utmost service as a means of investigating the In a drop of water, we are told, the number of molecules is so great that, if we could see them when the drop was magnified till it had, apparently, the dimenbe,
of the
minutest
structure of matter.
sions
of our
earth,
they
would appear
like
heap of tennis
balls in
Yet
in
number something like 9,556,000,000,000,000,000,000,000. each molecule we have two atoms of hydrogen and one
;
atom of oxygen, and, according to the latest views, within each atom of hydrogen there are some thousand electrified " corpuscles and it is the rhythmical vibrations of these infinitely minute executed perhaps five hundred million million times particles, every second, which we are able to investigate by means of the
spectroscope.
The
reality thus
globe of dew,
leaves
waken
In that
frail
With
To
will
These investigations into the infinitely great, and infinitely small, have for some minds the greatest attraction; others will,
* Sir
William Huggins.
Preface.
perhaps, attach
of
vii
more importance to the utilitarian applications of spectrum analysis and the services it has rendered in the discovery
new
in ordinary
and achievements of
its
analysis, which have marked There is a very large amount of work waiting to be done, both by those who have the command of fine instruments and elaborate apparatus, and also by those who have only the simplest means, and it is hoped that to these last, especially,
this
method of
latest
developments.
this
his thanks to
etc.,
many
friends
for
in
illustration
of this
spectrum analysis
to Sir
to to
the
Editors
W. Huggins;
Mr. Higgs,
spectrum,
obtained with
Rowland's grating ; to Professors Kayser, Runge, Paschen, Eder, and Valenta, and others.
W. M. W.
June
21, 1904.
CONTENTS
CHAI-TEE
I.
PAGE
How
TO PRODUCE A SPECTRUM
II.
Flame-spectra
Spectra produced by Means of Electricity
ii
III.
...
of
i8
IV.
....
34
V.
The
Diffraction
Spectrum.
Measurement
Wave-lengths
VI.
40
On the Production
of
Dark Lines
by Absorption.
58
75
VIII.
The
New
The
X.
Concave Spectrum
Photography
of
the
120
XI.
160
-
165
XIV.
174
185
Appendix
Index
....
311
321
AN INTRODUCTION
TO
HO IV TO PRODUCE A SPECTRUM.
Before we enter upon a description of the methods of spectroscopy, and of the various instruments as they are supplied by the optician, it will be well to consider what a spectrum is, and the simplest and most inexpensive method by which it can be produced. We shall thus gain some general idea of the subject at the outset, and it is hoped that the reader to whom
the subject
is
entirely
new may
We
must have,
in the first
parent material, with plane In default polished surfaces. of anything better, the lustre of a chandelier will answer our present purpose, such as shown If we place this in Fig. I. upright, and look through it in the proper direction at a flame
some distance off, we shall see Fig. I. the spectrum of a spectrum The best sort of the flame. flame for our purpose is a long narrow flame and non-luminous, such as the flame of a Bunsen gas-burner, or of a spirit-lamp, What we but if these are not at hand a candle-flame will do. shall see depends upon the nature of the light emitted by the
71
Introduction
to
the
yellovv flame. If the flame is non-luminous, and is then coloured by putting common salt into it, or by bringing a bead of melted sodium carbonate on the end of a thin platinum wire into it,^ we The shall see a yellow flame, but not where the flame actually is.
as rays of light from the flame are bent in passing through the prism, represents the position of the flame, and B shown in Fig. 2, where that of the observer's eye ; a yellow image of the flame is seen at Y.
Fig.
2.
Now, let a little lithium chloride be brought into the flame ; this colours it a brilliant red, and we see a red image of the flame at R, In like manner, some thallium less refracted than the yellow image. chloride colours the flame a beautiful green, and we see a green image of the flame at G, more refracted than the yellow image. These differently coloured and differently situated images of the flame constitute what we call the spectrum of the flame. shall improve our arrangements somewhat by putting in front of the flame a narrow vertical slit in a piece of cardboard or thin metal, so as to allow only a thin slice of light from the flame to fall upon the prism. The substances we have used to colour the flame were chosen because each of them gives only one colour, or, in other words, gives monochroinatic light but white light is a mixture of light of many different colours, so that an ordinary gas-flame placed behind the slit at gives us a succession of images of the slit of many colours without break, forming what is called a eontiimous spectrum red at one end, passing through orange, yellow, green of many tints into blue and violet. If we burn a short piece of magnesium ribbon behind the slit, we shall see a very brilliant continuous spectrum, with some bright green
We
lines
and
to the oxide
these are due to the vapour of the metal produced when the magnesium burns.
glasses, or pieces of coloured gelatine, or the sht illuminated by white light and
the prism, we get absorption spectra certain colours being cut out of the white light. Thus, if we put a test-tube containing a very dilute solution of potassium permanganate (Condy's fluid) in front of the slit, we shall see five dark bands in tlie green, due to the absorption of light of this particular colour by the solution, which must be only faintly pink.
:
How
to
Produce a Spectrum.
differences
o"o43 for
for
flint-glass,
and
this
is
is
said
greater
flint-glass has,
and consequently,
of flint-glass
and carbon disulphide of the same angle, the colours are more widely separated by the carbon disulphide than by the glass.
Let us now consider a httle closely the change in which takes place direction
more
when a ray of
liquely
Rl
to
represent the
path of
a ray of monochromatic light incident at I from air upon the surface If it suffered no change in direction it would have passed of water. into the water along IRi, but it will be found that it actually takes The angle RjIS, the change in direction, is called the the path IS. deviation, and is greater the less the angle between the incident ray and the surface of the water. Let us describe a circle QTS with I as
and S respectively ; centre, cutting the incident and refracted rays in refracting surface. the angles to right at normal) (the IQ draw also and SP upon the normal. and S let fall perpendiculars From
TU
An
I)itrodnctton to the
The law
of refraction
TTT
is
that
-^-=^
may
is
SIP
The
ratio
--^
TU
is
QIR;
is called the sine of the angle SIP. The refractive SI index, usually denoted by the symbol yn, is then the ratio of the sine of the angle of incidence to the sine of the angle of refraction ; for,
and the
ratio
SP
since
TI = SI,
/
V
TU
TI
SI
TU
'sp
The formula
want
If later on.
-.
=
r
or sin /
= u
sin
we
sin
shall, '
no doubt,
'
we look through a prism at a distant flame, as explained by and gently turn the prism round a little while watching the image, we shall observe that the image of the flame moves up to a limiting position Y (for the yellow light), and then moves away from this position again, although the prism is turned in the same direction
Fig.
2,
the time. This limiting position is the position of minimum deviation, and is the position in which the prism of a spectroscope is used.
all
Fig. 4
light
through a prism, at the angle of minimum deviation, when the course of the ray within the prism is equally inclined to the two polished surfaces. The angle of incidence at the first surface Def is equal to the angle
of emergence at the secoiid surface g/iY., and the deviation is twice the angle //cD,, and this angle is i r where
sin
Fig.
4.
ij.
sin r.
If the angle
is
of the prism
BAC
denoted
2r,
the deviation S
=
"f" ct
2 (/
;-)
=
a
2/
a.
by a, we have a Hence
and
S-f-a
z
= o-
and
-, so that 2
yu.
5iLL'
sin /
.
sin
sin
How
to
Produce a Spectrum.
This is the formula generally used in determining the refractive index of a substance, which is found by observing the minimum
deviation and the angle of the prism. We may use a simplified form of the above formula when the angle of the prism is only small. Since the sines of small angles are to one another nearly as the angles themselves, we may put i = /xr
instead of sin
i
=^
sin r, or
/n
-, or 8
(/a
i)a.
of parallel rays of light falls upon a converging lens, the rays are differently bent according to the distance from the central point of the lens upon which they fall, so that after refraction they are all collected approximately in one point, known as the focus. This is represented in Fig. 5, where F is the focus of the equi-
When
beam
Fig.
5.
(in this
Let us suppose that in Fig. 6, ABC represents a converging lens case a plano-convex lens), that O is the centre of curvature
Fig.
6.
sends out of the spherical surface, and that a luminous point at Let be one of rays of light which are refracted by the lens. these rays, which, travelhng along the axis QO, passes through the another ray which lens without any change of direction, and This ray is refracted as is refracted at B to meet the axis in q. if it passed through a prism whose angle is that between the plane and the tangent plane at B, which angle is equal to surface BOC. The deviation of this ray is equal to the sum of the angles ti, B^ z\ BO r, and BC a, since for small BQC, B^C. If we call angles the sines are nearly proportional to the angles themselves, we
Q QCO
QB
ABC
BQ
^
may
put -
+^
(m
i)
a
;
or -
i,r +-=
(ijl
j)
A/i Introduction
to tlic
The fact that a disappears from the equation shows that the distance of q from C does not depend upon the distance of B from and falling upon C ; in other words, that all rays emanating from the lens are collected at q, which point is called the focus of Q, or If we suppose to be very far off from the focus conjugate to Q. the lens, then the incident rays become parallel, and since u is now
mfimte, 7
ti
=
:
In
r
this
case q
is
nearer
the lens than is shown in the figure, putting F for the distance of the principal focus from of the lens the lens,
principal focus
the formula which F' connects the distances of the conjugate foci with the distance of the This formula is applicable to all principal focus from the lens. converging lenses, provided that the pencil of rays is a small one, so that no ray falls upon the lens very far from the axis. If the pencil of rays is not a small one, then, after refraction, all the rays will not go accurately to one point, and those furthest from This the axis will show the most deviation from the true focus.
we have
:;;
= (
-1
and -
+- =
'
7>
error
is
and
is
illustrated
by Fig.
7.
This
Fig.
defect may be remedied by the use of a stop, but, of course, at the cost of loss of light. There is in each lens a point called the centre of the lens which has this property, that any ray of light passing through the centre The direction of any such ray is called suffers no angular deviation. an axis of the lens. That axis in which the centres of curvature of the surfaces lie is called the principal axis, the others are termed secondary axes. In an equi-convex lens, such as shown in Fig. 8, the centre of the lens is the geometric centre ; in a plano-convex lens the centre lies upon the curved surface. All the rays of light starting from any one point of a luminous object, therefore, after passing through the lens, are collected into one point on that secondary axis which passes through the luminous point, at the distance given by
the
formula
t-.
^^'^
see,
therefore,
that
convex lens
gives an inverted image of the object, larger or smaller than the object, according as the distance of the image from the lens is greater
How
or less
to
Produce a Spcctrutn.
from the
lens.
7 This
is
than
the
illustrated in Fig. 8.
What has been said about lenses is true for light of any one colour or monochromatic light, but when using white light the image will be found to be coloured, and this defect of a single lens is termed " chromatic aberration." Since violet light is more refrangible than red, there will be formed a violet image somewhat nearer to the lens than the red image. This defect can be partially remedied by the use of an achromatic lens, made of two lenses cemented together ; a convex lens of crown-glass, and a concave lens of smaller curvature made of flint-glass. The possibility of constructing an achromatic lens depends upon the fact that the dispersive power of flint glass is greater than that of crown-glass in proportion to the deviation pro-
FiG.
8.
if we have two prisms, one of crown-glass and the other of flint-glass, of such angles that they produce the same deviation of a green ray, the red and blue rays will be more widely separated by the flint-glass prism than by the crown-glass prism. This will be seen from the following table of the refractive indices of flint- and crown-glass for light of difterent colours
duced, so that
An
Introduction
to
the
Fig. 9 represents an achromatic lens, and Fig. lo a direct-vision prism composed of three prisms of crown-glass and two prisms of
flint-glass.
employed
for chemical
represented in Fig. ii. The light from the flame e passes through a fine slit in the plate at the end of the tube A, which is called the collimator-tube, through the prism P, and into the observing telescope B. An
analysis,
is
white
tiff
At
I-"IG.
9.
Fig. 10.
I'lG.
II.
ment required
is
slit-plate is given in Fig. 12. The adjustthat parallel rays should pass through the prism
Fig. 12.
at the
angle of
minmium
deviation.
The
collimator
is
furnished
How
to
Produce a Spectrum.
with a convex lens at the end nearest the prism, and the slit is adjusted at the principal focus of this lens, so that the rays fall upon the prism as if they came from a slit at an infinite distance. To put the instrument into adjustment, the observing telescope should first be removed and focussed exactly on a distant object it is then screwed into its place, and the arm which carries it turned round until the telescope is in a line with the collimator ; the slit-tube is then drawn out until it is seen in sharp focus on looking through the telescope, and the adjusting screws a and /3 are moved until the middle of the slit is seen in the middle of the field of view. The prism is then placed on the table, and fixed in the position marked for it by means of the clamp y. This position is that of minimum deviation for the green rays, but it is not necessary to alter the prism for the red or blue portions of the spectrum. The third tube, C, carries a photographed scale and a lens at the end near the prism, similar to the collimator. The next adjustment is to get this scale (which is seen by reflection in the face of the prism) in the right place and in sharp focus. The arm carrying this tube, turns round the central pillar, and can be fixed in the right position by the set-screw /i ; the scale can be raised or lowered by means of the screw 8 ; the tube can be pushed in or out to focus it
It is usual to make the image of the fixed edge of the slit coincident with the number 50 of the scale, the scal& being illuminated by a small luminous gas-flame placed behind it. The use of the small movable prism shown in Fig. 12, which covers up half of the slit, The little is to enable the observer to directly compare two spectra. prism sends the light of the flame D through the slit by total reflection, whilst the light from the flame E passes through the upper portion of the slit directly. It is not possible to construct scales for different instruments which shall give exact/}' the same readings, and of course the scale is an arbitrary one. Bunsen's maps are given on the scale of his own spectroscope, in which the sodium line is placed at 50, and the lithium, strontium, and thallium lines read as follows
exactly.
67-8.
of that edge of the sht which does not move when the altered is made to coincide with the division 50. If this be on the left of the observer, then always the position of the left-hand edge of each line and band is to bf recorded, and in the case of a faint line the slit may be opened to admit more light, and This refers, of course, yet an accurate reading may be obtained. only to lines which are sharply defined, and not to bands of con-
The image
width of the
slit is
siderable breadth.
plan in making a map of a spectrum is first down, as exactly as possible, the positions of the well-defined lines on a millimetre scale on paper, opening and closing the sht as convenient, and then to go over the work again, keeping the slit at one uniform width, and noting the relative intensity of the lines, and
to put
10
An
Introduction
to
the width and character of the bands, whether sharply defined at the edges, or sharp at one edge and fading away at the other, or bright in the middle and fading away at both edges. There is no better plan of recording the peculiarities of a spectrum than that employed
CHAPTER
The
II.
FLAME-SPKCTRA.
spectrum of an incandescent solid is continuous ; a discontinuous spectrum of briglit lines is produced only by an incandescent gas. The yellow line seen when a salt of sodium is introduced on a fine platinum wire into the non-luminous flame of the Bunsen burner is due to the vapour of sodium set free by the high temperature of the flame, or by the chemical changes taking place in the flame between the hot gases of the flame and the salt introduced. Bimsen remarks that all sodium salts give the same reaction ; although it is obtained more vividly from the more volatile compounds, such as the oxide, chloride, bromide, iodide, carbonate and sulphate, yet the non-volatile phosphate, silicate, and borate show the line plainly. This reaction is of extraordinary delicacy Bunsen showed that o"oooooo3 milligram of sodium could be detected with certainty by the spectral reaction. Indeed, the reaction is too delicate for the comfort of the spectroscopist, who finds the sodium line almost always present, even when no sodium compound has been intentionally brought into the flame. Spectrum analysis thus reveals the wide diffusion of substances previously accounted rare ; as of lithium in sea-water, river and mineral waters, in tea, coffee and tobacco, and The invention of spectrum analysis was quickly in human blood. followed by the discovery of new elements ; of rubidium and cassium by Bunsen, in the mineral water of Durkheim ; of thallium by Crookes, in 1861 ; of indium by Reich and Richter, in 1864; of gallium by Lecoq de Boisbaudran, in 1875; ^"^ "O'*^' o^ '^^ "'^"' gases, argon,
:
helium, krypton, neon and xenon. In Fig. 14 we have shown a large spectroscope of nine prisms, constructed by Browning for Mr. Gassiot, known as the Kew Fig. 15 shows the path of the rays of light through Spectroscope. represents the collimator, seen on the left in Fig. 14, the prisms; and B represents the observing telescope. The carbonates, chlorides, or nitrates of the alkalies and alkaline if they can be earths are convenient salts to use ; but the chlorates obtained are perhaps to be preferred, as in consequence of the
oxygen the flame is hotter, and the reaction more In a flame of low temperature the reaction is much less distinct; thus in the flame of hydrogen burning in chlorine the
liberation of
brilliant.
12
/I
hardly seen is not obtained, and the yellow line is use of a The air. in burning hydrogen sulphuretted of flame the in hotter flame may bring out new lines; thus in the oxy-hydrogen
sodium reaction
-*
flame sodium
salts
line,
a series of
and
83'6.
Lithium
salts
'\
Flame-Spectra.
13
of which consists of two lines, a very brilliant one at 31 '7, and a much feebler line at 45'o. If the oxy-hydrogen flame be employed, there is seen in addition a blue line at io5'2 close to, but a little more refrangible than, the strontium blue line. All lithium compounds show the red line distinctly, even minerals containing lithium such as lepidolite only need to be held in the flame to give the reaction
distinctly.
The spectrum of potassium is remarkable, consisting of a line in the extreme red and one in the extreme violet, at i7'5 and i53'o respectively, besides a certain amount of continuous light. Other lines come out in the oxy-hydrogen flame at 51 "8,
52-4, 53-2, 67-1,
and
i52;7.
Some
of the silicates containing potassium do not show the potassium lines until they are
;
heated with sodium carbonate the potassium felspar " orthoclase "
is
in
this
way
easily
The
spectra of csesium
and
ends of the spectrum, and all three showing a continuous spectrum in the centre. Rubidium (from ptijStSos, dark-red)
is
Fig.
named in allusion to the red lines at 15 5 and i3'5, even less refrangible than the potassium red line; it has also characteristic lines in the violet at i3S'5 and i37'o, and several less characteristic Csesium gets its name from the bright lines in the yellow and green. blue lines at 106 and 109 (from casiiis, sky-blue) ; the spectrum has several other lines in the red, yellow, and green, the brightest of
which is at 42. Like potassium, rubidium and cassium are precipitated by platmum
the quantity of these rare substances in the substance is small, the spectrum test is best made with the precipitate precipitate obtained with platinum chloride, by heating this with a few drops of in a current of hydrogen and boiling the residue
chloride
if
under examination
water.
The spectra of the alkaline earths, barmm, strontmm, When first calcium, are not quite so simple as those of the alkalies. which are bands certain seen are there flame the into introduced metal used, and which different according to the particular salt of the employed ; but the final are supposed to be due to the compound
and
14
An
Introduction
to
the
salts of any one is the same with all the different produced by the metal, and is partly due to the oxide of the metal there appear the reactions taking place in the flame, besides which the strontmm brightest lines due to the metal itself; for example, due to the blue line at 105, and the calcium violet line at 135, are The lines attributed to the special compounds, e.g;. chloride, metal. bromide, and iodides of these metals, are maintained and strengthened introducing into the flame a current of hydrochloric acid or
spectrum obtained
by bromine vapour, or iodine vapour, as the case may be. Tltfese differences are clearly shown in Fig. 1 6, copied (by permission) from
the drawings by M. Lecoq de Boisbaudran, given in his valuable work " Spectres Lumineux," on a scale a little different from Bunsen's The spectrum of calcium is at once recognized by the scale. characteristic orange band, 40 to 43, and the green band, 61 to 63. The chloride, chlorate, bromide, and iodide give the reaction best the non-volatile salts must be treated with hydrochloric acid (if it
decomposes them) or heated with ammonium fluoride. The strontium spectrum is characterized by the absence of green lines; the brightest lines are the orange band at 44 to 47, and the red lines at 30 and 35, but the most characteristic line is the blue one The reaction is given best by the chlorate or chloride or at 107 '6. but it is not given by the sulphate, phosphate, siliother haloid salt cate, etc. These salts require a preliminary treatment. The sulphate may be reduced to sulphide by heating in the reducing flame of the Bunsen, and then moistening the bead with hydrochloric acid the silicate or phosphate should be fused with sodium carbonate in a the residue, after being washed with little spiral of platinum wire hot water, is moistened with hydrochloric acid, and brought into the
;
flame.
is
presents a splendid series of green bands, the The non-volatile salts of brightest of which are at 73 and 78. barium require the same treatment as those of strontium. In Fig. 16 the lines due to the metal are marked M, and those due to the oxide are marked O. In the Bunsen-flame spectrum of
barium chloride there are three metal lines, 440, 6f6, and 69'9 the strontium spectram shows two metal lines, 477 and ro7'6; whilst the calcium spectrum shows only one line, at i38'2, due to the metal. The lines due to the compound of the metal with chlorine are 69'8, 72*9, 74'S, 76'o, 77'6, and Si'o; the rest of the lines are no doubt due to the oxide. In the strontium spectrum the lines due to the chloride are 30*3, 33'o, and 38'25. In the calcium spectrum the lines due to the chloride are 4r'75, 42'25, 45'o, 45'7, 488, 52'2 whilst 6i'2 and 62 ! seem to be due to the oxide. We may therefore conclude, from spectroscopic evidence, that calcium has a greater affinity for chlorine than strontium and barium have, and that barium has a greater affinity for oxygen than strontium and calcium have. Other metals (besides those of the alkalies and alkaline earths)
; ;
Flame-Spectra.
15
^^
^. .^.J V^
|
lin' liii
l
| |
Barium. Chlortde,
iiii|iiii
|
Ftame
[Ull|llli| ilTT]Tll11lfrffl lf
nii
|
ii ii|
iiii
|
iiiim
20
30
40
50
60
mi
| |
70
80
Wi
| |
90
100
lie
120
130
140
ISO
160
Baruan
i|iiii
|
Chloride,
iiii
| |
Flame
iiii iiii
| |
chortled -with
iiii
| |
HCl
iminn
iiii|iiii|iiii iiiiniii|iiii
|
iiii|iiTi'
iiii
|
Hii
iiii
|
iiii
iiii
|
iiii
|
20
30
40
SO
60
70
80
90
100
no
120
130
140
150
160
iiii
iiii
iiii
|
iiii
|
iiii
|
i!ii
|
iiii
|
iiii
|
20
30
40
SO
60
70
60
90
100
110
120
130
140
150
160
Barium
iiii
|
iiii
|
iiii|Nii
|
iiii|iiiT iiii
| |
ini
|
iiii
|
iHiiiiii
|
iiii
|
iiii
!iii
|
iiii
|iiii|iiii|iNi
mriTT^
I
2Sr
30
S>
40
so
IIMO
60
80
90
100
110
120
130
140
ISO
160
ini
iiii
niimii|iii[|[ii[|iiii
iiii
iiii
iiii
80
90
100
no
120
130
140
ISO
160
StrontUtm, Chloride.
FlcLTThe
llll
[
with
HCl
llll
|
llll
|
llll
|
llll
|
llll
|
llll
|llli|IIIITTTTTjTTrn
80
90
100
no
120
130
140
150
160
Cahxum
i
Chloride. Flcune.,
iiii|iiii
| | |
nii|iiimiii|Nii
| [
iii| iiii[iiii
iiii
iiii
|
iiii
[
iiii[iiii
[
iiii
|
iiii
|
80
90
100
no
120
130
140
150
160
iiii
|
iii!
|
irTTpTrini|ini|iiiimii
|
iiii
|
iifi
|
iiii[iniiiiii
|
iim
20
30
40
SO
60
80
90
100
no
120
130
140
150
160
CarboTv
(Bltnseix.
|[ni|i[ii iiii|iiii
| [
Ftume.)
i[1t 1in[iiii
| |
nltth1i[iiii
[
iii[
|
mTiiii|iiii i[ii|iiiipi
| |
iiii
|
iiii
[
20
30
40
50
60
70
80
90
100
no
120
130
140
150
160
SodivLTn,
iiii|ini
|
spark.
iiii[iiii iiil!|iiii
iiii
|
ilii|iin|iiii|iiii
|
ilii
|
iiii[iiii
|
iiiijnli
|
nli
|
iiii
|
iiifpi^^
120 130
140
150 160
20
30
Li
40
50
"'^
60
70
80
90
100
no
n n
Lj/Juum,
in.
the
Flame
iiiipT[frm
ini
|
lllip[TTTr
Nii
|
iNi
|
iiii]iiiniii[
[
20
30
40
SO
60
70
80
90
100
no
120
130
140
150
160
Fig. 16.
An
Introduction
to
the
J
iii
l
^.
[|
l
^
il
iiiilifli
illl^lfll| n
ll
l|il
lfllttp
|ifll
[
iiii
il
llft|llli
|
p
80
!iTfni |llll
|
ltt^
^^
120
Copper Oxide
20
30
Copper
40
Chloride.
50
60
70
90
100
no
130
140
ISO
160
iiii|[iii
| |
iiii|iiiimii ini|iiii
| |
iiyi |ifiijiii[l1ll1flll]rllf[lfir]1fliniTT
60
70
i1[i
|
ii1^[nif|l1ii|tifi|ii1l]iiliniii
|
iiii
|
iiifpTn|
20
30
40
SO
80
90
100
110
120
130
140
ISO
160
Gold, Cklvrid&
iHi
|
iiii|iiii|iiii
|
im
50
20
30
40
T 60
60
ltTl|lllllllll|llll|l1T
iiii
|
iiii|rfmTnTi
70
80
90
100
110
120
130
140
ISO
160
J^iarigaiwse Chlorideiin|iiii
| |
iin
|
iiir
i[i| iiii
|
iiii|i
20
30
40
50
70
80
90
100
no
120
130
140
150
160
[TM]]T]
BoncAcid,
1
1111
1111
III!
lin| ITTT[lll[|llll
[
llll
[
llll
[
IIN[llll
|
llll|IITTpi
20
30
40
SO
60
70
80
90
100
110
120
130
140
ISO
ISO
Phosphurette-d Hydrocfen.
ii
|
iiiiniii
|
iiii
|
iiii
|inTTiii[
iiT iiii|ini|1ffl
[ [
iiiTpin iiirpii
[ |
iiii
|
iiii
|
iHi
|
iiii
iiiimii|iiiiiiiii
|
iiii
|
iiii| iiiriiiii|
20
30
40
50
60
70
80
90
100
110
120
130
140
ISO
160
Air.
nii|iiii|i'ili]!ill
|
+poteJ
h^
|
'liit[iii^iiihiit[iin^^
20
30
40
SO
60
70
80
90
100
110
120
IJ" 130
' 140
"" ISO
" 160
Air, I -pole)
i1ti| fii'
i [
iii
i]
iiii ff ift|tfri
M|
iii rriiii'
iiM
100
il
[rri
i|
bi
trr
|
m
130
20
30
Jl
40
50
N i
ii
60 2L
,1
70 o
80
N N
90
110
120
ItO
ISO
160
o
..I
o onS
nno
Air,
longsparlc.
ni[[nN
|
i
|
liif^
110
ii
i
mi iiiiiii
140
ISO
160
20
30
40
50
60
VI
70
CI
80
a,
90
100
120
130
H
^
a a
H
Air, short spark.
'iiMilffl^
| I
|llii|iiii| iiii|iiii
|
iiif iiii|iiii
| |
illipil'
| I
Mii[pifiiil
I/*
I
20
30
40
50
II
60
L 70
n^
130
80
90
100
110
120
140
150
160
i|ii li
|
't
nft
ii
|
k i
ii]
|ii'ff[Nit|mt|
iii
[
iiil
p^
^
90
100
110
20
30
40
50
60
70
80
120
130
140
150
160
Fig. 17.
Flame-Spectra.
do not, as a rule, give spectra when their salts are brought into the Bunsen flame, either because they are not volatile enough, or because they are not decomposed at the temperature of the flame and for these it is necessary to employ a higher temperature by the use of electricity. There are, however, a few substances which give flamespectra,, and these may conveniently be described here before we proceed to describe the various ways of using the electric discharge. Copper chloride colours the flame green or blue, according to circumstances a bead of the substance brought into the flame melts and gives a yellow patch of light bordered with a reddish tint, outside this central colour is seen a blue colour, and beyond this, on the
; ;
outside of the flame, a green colour. On continued heating, the chloride of copper becomes converted into oxide, the blue colour disappears, but the green persists j on bringing hydrochloric acid into the flame the blue colour is restored. Professor Smithells has shown that the green colour is due to an oxide of copper, the blue colom- to cuprous chloride, and the yellowish-red coloured portion of the flame is due to metallic copper. The spectra of oxide of copper and chloride of copper are shown in Fig. 1 7 ; the two spectra are generally seen together. Gold chloride gives a greenish colour to the flame, the spectrum of which is drawn in Fig. 17. Hydrochloric acid is essential to its production, and the spectrum is therefore regarded as due to the compound of gold and chlorine. Lead chloride colours the flame blue, and behaves very much as copper chloride does, giving a spectrum due to oxide of lead, besides lines due to the chloride. Bismuth chloride gives a spectrum of bands due to the oxide, and on continued heating a line at 100 (wave-length, 47227) due to the
metal.
Manganese chloride and manganese oxide give a fine spectrum of bands due to the oxide, especially in the oxy-hydrogen flame.
This spectrum is drawn in Fig. 17. Other substances giving flame-spectra are boracic acid and phosphuretted hydrogen their spectra are also given in Fig. 1 7 A trace of a phosphorus compound brought into a hydrogen-flask sufifices to colour the flame of the hydrogen green, and to give the three green lines at 73, 79, and 59.
;
CHAPTER
In
order
III.
Fig. i8.
least a
two-inch spark
in air when worked by four or five Grove cells, number of bichromate cells, if these are found more The spark may be made to pass between a piece of the
metal under examination in the form of wire, or pieces held in the forceps of a spark-discharger as shown in Fig. 19. Sometimes a Leyden jar is used, which is connected as shown in Fig. 20. The spark between metal electrodes gives both the spectrum of the metal and also that of the air or other gas surrounding the spark. To distinguish the lines due to the metal from those due to the air, the spark may be surrounded by hydrogen ; it is for this purpose that the httle glass tube shown on the right in Fig. 19 is used. The lines
spectra produced by
Means of
Electricity.
19
due to hydrogen are few in number and easily recognized, whereas those produced by air are exceedingly numerous, and may all but obscure the lines due to the metal. The relative extent to which
Fig. 19.
the lines due to these two sources appear depends very much upon the volatility of the metal, and the length of the spark. If the spark is only weak, the metal lines may be seen only in the close neighbourhood of the poles, and are thus discontinuous, whilst the air Hnes
I-'IG.
20.
{From
go
The appearance of the right across from one pole to the other. spark may be altered by regulating the adjustment of the contactWhen this is made to work more slowly by breaker of the coil. slackening the tension, the metal lines become brighter, and those of
20
An
Ititroduction to the
perhaps, still better. It is, therefore, very important to make oneself .familiar with this spectrum of air, and it is therefore given, under different conditions, in Fig. 17, where No. 7 shows the spectrum of the positive pole with a spark of moderate length (the band spectrum of nitrogen); No. 8, the spectrum of the negative pole under the same conditions ; No. 9, the spectrum of a long spark (lines of oxygen, nitrogen, and hydrogen) and No. 10, the spectrum of a very short spark between
is,
;
In
this last
draw-
ing there are shown lines of hydrogen at 33'7, 9175 ; lines due to chlorine at 64-25, 6472, 65-58, 66-80, 74-00, 79-18, 80-30, 88-47, 89-67, 94-30, 95-47, 96-05, and 96-80; and lines due to platinum at 63'SS and 70-38. There is sometimes an advantage in using a lens between the
FiG. 21.
source of light and the slit ; this is adjusted so as to throw a sharp image of the source of light flame or spark upon the slit. By this means we can distinguish the different spectra proceeding from different portions of the spark or flame. If the experiment be tried with an ordinary candle-flame, using a small hand-spectroscope with the slit horizontal, we get, when the middle of the image falls across the slit, the continuous spectrum of white light ; but if the very tip of the flame falls on the slit, we have the one bright yellow line of sodium, and if we make the image of the blue base of the flame fall upon the slit, then we get the spectrum of bright lines (to be afterwards described) known as the " Swan spectrum. If the spark is placed horizontally, and the image of it thrown on
to a vertical slit, the spectrum of the vapour surrounding each pole can be more clearly distinguished from that of the surrounding gas. With this arrangement all the lines of any one substance are seen not to be of the same length. This is shown in Fig. 21, which is copied from a photograph of a portion of the spectrum obtained from the
spectra produced by
Means of
Electricity.
arc between poles of carbon impregnated with a strontium salt which contained calcium as an impurity. The longest lines at 4078 and 4215 are due to strontium; there is a long line of calcium at 4227, and the four short lines on the left are due to impurities of calcium
in the carbon poles. These four lines two outer ones are due to calcium, and the two inner ones to aluminium) are always seen in the photographs of arc spectra, whatever the substance used. We get a hotter spark and brighter spectra by using a Leyden jar, of which the inner coating is connected to one terminal of the coil, and the outer coating to the other terminal. This arrangement is shown in Fig. 20; in Fig. 19 the Leyden jar is contained in the box which forms the base of the instrument.
in-
FlG. 22
Fig. 22.*
form or solutions of purpose are shown in Figs. 22 and 22*. The apparatus shown in Fig. 22 is for the use of dry salts. It consists of a little conical cup made out of an aluminium rod about \ inch in diameter; a fine hole is drilled in the bottom of this cup, into which a platinum wire is inserted (like the wick of a candle) ; this cup is employed in the spark-holder as the bottom electrode, the upper electrode being a piece of wire of copper or some other metal ; the salt to be examined is pressed into the cup, the use of the platinum wire being to confine the spark to the centre of the salt, and thus to render it steady. The apparatus shown in Fig. 22'"' is for use with solutions it is made out of a piece of glass tubing, about \ inch in diameter, closed A platinum wire is melted through at the bottom like a test-tube. the bottom to serve as the lower electrode, and a bit of thin glass
vestigated,
its salts
we may employ
in the solid
the
salts.
Convenient arrangements
for this
22
An
Introduction
to
the
;
tubing is dropped over this the solution to be examined is put into the tube and adjusted to such a height that it rises by capillary The upper electrode is a attraction to just cover the electrode. platinum wire, melted into a glass tube which jiasses through a cork fitting the top of the test-tube, and which is easily adjusted within the The two other glass required distance from the bottom electrode. tubes shown, passing through the cork, enable a current of hydrogen or
other gas to be passed through the tube to displace the air, if required. The work of M. Lecoq de Boisbaudran, entitled "Spectres Lumineux," contains a very careful series of measurements of the spark spectra of the metals, as seen in a short simple spark in air, between platinum poles moistened with the solutions of the metallic These are accompanied by an atlas of drawings, from which salts. many of our figures are copied by permission. M. Lecoq de Boisbaudran employed a coil giving a spark of from 2 to 5 cm. length he recommends the use of a rather short spark with solutions of the chlorides of the metals ; he finds that, under these conditions, the lines due to the air and to the electrodes do not interfere with
the results to any great extent. On comparing the drawings of the spark spectra of sodium, lithium, and potassium with those of the flame spectra of the same elements, we see that the higher temperature of the spark gives rise to many more lines so that, for example, instead of the one line of sodium and the two lines of lithium in the flame, we have in the spark eight lines due to sodium and four lines due to lithium, and thirty-one lines in the case of potassium. Some of these additional lines are obtained in the flame spectrum by feeding it with oxygen instead of
:
air.
to
It seems, therefore, as if these additional lines were due simply With barium, the higher temperature of the source of light. strontium, and calcium the case is somewhat different, since in the
flame spectra we have bands due to the oxide of the metal and to the chloride or other salt used, and some of these bands disappear at the higher temperature and are replaced by additional lines due to the metal itself. In the spectrum of manganese chloride produced in the gas, we have a series of bands due to the oxide of manganese. In the spark, not too short, we have these bands again, as well as certain conspicuous lines due to the metal itself; and if we make the spark short enough, the bands of the oxide disappear, and the lines of metal become more numerous and decided. These spectra of manganese are shown in Fig 17, No. 4. We get also a spectrum of bands due to the oxide with aluminium, if we employ electrodes of the metal, or if we use aluminium chloride at the negative pole. The bands of copper chloride, shown in Fig. 17, are obtained, together with the lines due to copper itself, in the spark with copper
chloride.
in Fig. 23.
with
magnesium chloride
is
shown
shows a very strong double line at 76, and a shaded band beginning at 83-8," and fading away towards the blue end of the
*
is
shown
at
SS'S
liy
mistake.
spectra produced by
Means of
Electricity.
23
Chloncfc
^an^anese
MTTTI m\\\\\
I
short
iijfii
I
spofh
\mss^
SO
' '
MJJi llMiin|niin]ii^Lmiijl|ttljiiJi^
20
30
40
60
70
I
80
I
90
100
NO
120
130
140
ISO
160
|lili|liil|illl|llll|lllipjllf^^^
20
30
40
50
60
70
80
90
100
110
120
130
140
ISO
160
^Ma^nsiianCfdoride, spark.
fT1TI|TTT1 |llll[II
H |llll|lll!i||||| n
lll
[
ll ll
[
llll[T[im^
20
30
40
50
60
70
80
90
100
110
120
130
140
ISO
160
24
Alt Introduction
to
the
spectrum. This band is due to the oxide of magnesium, and is obtained very brightly in the spectrum of burning magnesium together This with the double line at 76, which is due to the metal itself. line is really triple, and can be seen as such if the slit be made very " little b " in the narrow. It coincides with the dark line known as Three other less conspicuous lines of the metal are sun's spectrum. seen at 61-7, 101, and ii6-6. These are the chief examples of oxide spectra produced in the spark ; in all other cases the lines seen in In the spectrum of zinc we spark spectra are those of the metal. have three bright lines at 94-6, 100-05, and 102-6, forming a wide triplet, and in the spectrum of the allied metal, cadmium, we have In the spectrum three similarly spaced lines at 80, 95-4, and 102-8. of silver the chief lines are two green ones at 64 and 74-4. Mercury and mercury salts give a spectrum in which the chief lines are a yellow double line at 53-0 and 53-7, a green line at 64-1, almost in the same place as one of the silver lines, and a blue line at 126-4. This spectrum may also be obtained from a vacuum-tube containing mercury highly exhausted, and is given very brilliantly by the Cooper-Hewitt mercury-light, where the arc-light is produced
between mercury electrodes. In the spark spectrum of lead the chief line is a blue one at 155-9. If a solution of lead chloride is employed, this is the only line seen, but with a solution of the nitrate we have also the lines at 84, 74-7, The spectrum as drawn was obtained by the use of 59, and 46-8.
electrodes of metallic lead. Salts of thallium in the flame give one brilliant green line at 68-4, and a second faint line at 56-5. When the spark spectrum is examined, besides the flame-line in the green there are seen two hnes in the red and orange, several lines in the green, and a well-defined line in
the blue. means of
in
1861 by Sir
W. Crookes by
spectrum.
In 1864, Drs. Reich and Richter obser^'ed two deep-blue lines in the spectrum of a variety of zinc ore, and were thus led to the discovery of a new metal, which received the name " indium." Indium chloride colours the flame deep blue, and the spectrum shows two lines in the blue at 114-5 and 150-8. The spark spectrum of tin is characterized by two lines, a blue
line at
113-4, and a yellow line at 58-2; two lines a little more refrangible than this yellow line, at 59-6 and 60-6, are feeble, but if a
Leyden
jar
is
in the
whole
specti-um.
The most
60-9, 74-3,
^
and 149.
Antimony only gives a bright spectrum with strong solutions ; the chief lines are in the red and greenish yellow at 46-8, 60-3, 43-5,
and
44-8.
A
and
brilliant
solution of chloride of gold subjected to the spark gives a spectrum, containing some of the bands of the flame spectrum, in addition, sharp lines due to the metal. The most characteristic
spectra produced by
Means
of Electricity.
25
Bismuth. spark
[
iii'|i
H |Hi
i
iiii
| [
if
l|
iii
|1
|n
i|
1iii
|
iii
mn
70
'
f
|
f fi
iTii
| |
riii|ii
ijTTTf iiii
|
iiii
|
iiii|iinmir|TTTTrfr[i|nTT[iiii iiilmii|[[ii
| |
20
30
40
SO
60
80
90
100
110
120
130
nO
ISO
160
uiii
l
iiii
iili
ii
JllMljJ
50 60
70
20
30
40
80
90
100
110
120
130
140
150
160
l1 ii
|iln |ii
it]
iFTT
|
^
100
110
I 1 1
iii| iili(iiii
|
iiii|nii[iiii
|
iiii|iiii|
20
30
40
SO
60
70
80
90
120
130
140
150
160
Platirutnh,
I
ll
sparh
P|lIil|M'|llll|j:UJ |ff ll
[
'Nl[l
[l |l
lfr[l11ip
|^^
20
30
40
so
60
70
80
90
100
110
120
130
140
ISO
160
ill
|iiii
|
ll
[iil
II
III
^'^^^^^/^^^^^^'
1
I
iii
l[
ii
|
ii
[iM
|
i|iiii |lfff|lll
ipl|^^
20
30
40
SO
60
70
80
90
100
110
120
130
140
150
160
[iiii|iiii|iiii|ii[i[llif|riiki[Uif|'iiil|fllf|'^^^^^^
Ai,lHilJli..,JilJiii
II
\UL\JZ";-JP'':L
120
130 140 ISO
160
20
30
40
60
60
70
80
90
100
110
Cohalt, sparh.
ii ii
| | | [
|
iiii
i|
iii
ii
i|irii
iiii
|
iiii|iiff|iiti
|W| nn
i
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
|llll|llll|lin| llll|'lfll|llll|lliiyil'l|[TTTTT]TI^
20
30
40
50
60
70
80
90
100
110
120
130
140
ISO
160
IctajlJUTTV.
|lll
iiiHm
20
30
40
50
60
70
80
90
100
110
120
130
140
ISO
160
26
An
Introduction
to
the
of these are a yellow line at si'6 a"d a red line at 39-9, then green and blue lines at 957, 73'4, Si, and STiThe spark between electrodes of platinum does not give lines due to the metal unless the spark be very powerful or a Leyden jar be employed; with the ordinary spark only air lines are seen; but if the electrodes are moistened with a solution of chloride of platinum It is advisable to a brilliant spectrum of fine lines is obtained.
employ a short spark and moderately concentrated solution. The chief lines are 63-6 and 70-3, then 8r-25, 73-5, and 66-8. With a solution of chloride of palladium the chief lines are green lines at 707, 76-3, and 6i'r. They Iron, cobalt, and nickel give very complicated spectra.
are best obtained from solutions of the chlorides of the metals. The most characteristic lines of iron are three green lines at 7i'8, 69-3, and 73'4, then two blue lines at 86-3 and 88-25, and a violet
line at i24'3.
68'25 and
The most characteristic lines of cobalt are a fine double line at 6875 in the green, a somewhat wider double line at 7i'3
7i'9, a single line at 74'2,
and
and blue
lines at 9r'3
and
92-9.
characteristic lines of nickel are a bright green line at 63"6, another at 8o'2, and a blue line at 100-4. The most characteristic lines of titanium are 40-4 in the red, 62'2,
74'3, 8o'9 in the green,
The most
lines, iio'i,
1107, iii'b,
and 1127.
Molybdenum
at 60-3,
is
recognized by a group of three lines in the green and lines in the yellow and orange at so'i
;
and
46'i.
The simple spark between fragments of arsenic gives only a continuous spectmm, but if a Leyden jar be employed, a very large number of lines are obtained. Arsenic may, then, be recognized in minerals containing it by red lines at 42-5 and 44, green lines at S7'S, 607, 62-8, 69'2, and others. Only the strongest lines are shown in
the figure.
is recognized by the blue line at no; minerals conmust be fused with an alkaline carbonate. Count de Gramont, in his treatise on the application of the spectroscope to the
Beryllium
it
taining
recommends carbonate of lithium ; the mineral fused with the lithium carbonate in a loop of platinum wire placed in a Bunsen burner, and the spark is made to play upon it while in the flame, so that the bead is kept liquid or
analysis of minerals,
in
fine
powder
is
pasty.
The
method.
silicon
in
silicates
The most
red at 47-8 and the double blue band at 147 to 148 is also very characteristic, especially if the method of photographing the spectrum be employed. Germanium is recognized by two broad and bright hues in the orange at 46-3 and 50 this last line is apparently coincident with the sodium line, but it is really a little more refrangible.
;
characteristic lines of silicon are two lines in the 48-4, and a double green line at 81-2 and 8r9;
spectra produced by
Means of
Electricity.
27
Ar$enic.
iiii| iiii|iiii
i
iiii
iiil
|
iifi
|
iiilliiii
niriiii
[
iiii|iiiiiiiii
|
iiii
|
iiii
|
iiin
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
BeryllifUtv.
D il
i
i[
ii[iiii|iiii|
iii|ir[Tpii[iiii|ii]j.|Jin|
20
30
40
SO
60
70
80
90
100
110
120
130
MO
150
I6&
Silicon.
iii|i iii
|
ini
[
ii"i
[
iiii|iili
|
iiii
|
iiii
|
iiii[iiiijiiii
|
iiii[nii
|
iiii
|
iiii|iiii
|
iiii
|
iiiifilii
|
iiii[iiii|iiiim
20
30
40
50
60
70
80
90
100
no
120
130
140
150
160
OernvattiitTn.
iiii
|
iiii
| |
iin nii|ni[|iiii|mi
[
iiii|iiii
|
iiii|iiii
|
nii
[
iiii
|
iiii|iiii
|
iiii|iiii|iiii|iiii|iiii
|
iiii
|
iiiim^
20
30
40
SO
60
70
80
90
100
110
120
130
140
150
160
Van.a,(tia7n.
iiii|iiii|iiii|iiN
|
|
iiii|iiii|iiii
|
Tm|TFi|iiii
|
iiii
|
iiiim^
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
ZirconiuTTu
iiniiiii|iiii|iiii[iiii
|
iin[ini
[
iiii|il[i
|
iiii[iiii
|
!iii|iiii|iiif|ili]i[ii iiii
| |
iiii|iiii|iiii
[
ii^
20
30
40
SO
60
70
80
90
100
no
120
130
140
150
160
iiii|iiii
|
|
iiii|iiii
|
iiii|iiii
|
iiii|iiii|iiir
iiTiiinii[i|iiiimii[iiii|iin iiii|iiii|iiii|iiimiii
| |
jd
70
Chromiwnh.
fi1i
i
i[fi
iiii
iiii
iiiiiiiiiirrrr
20
30
40
SO
60
80
90
100
110
120
130
140
150
160
\Carborh. spark.
Ft
20
iii
iin
ini|!iii|iiiimii|inimii|ihimii|ini|iiii|iiii|iiii|iiii|iiii|iiii|iin|iiii iiii|iiin^^
[
30
40
50
60
70
80
90
100
110
120
130
140
ISO
160
iiii
| |
ini
|
iiii
LVI
i|'iiii
|
ii
Seleniicm.
f l
|
i|iiii
'ln
|
i f
fi lli y
ii[ii!i
|
nV
i |
iiii
^^^
20
30
50
60
70
80
90
100
no
120
130
140
150
160
Teliuriu-ftv.
|iiii
|
iiii
|
iiiniiifrr
wyf[MT^m
|
l^ l
l| flfl
|
llll
|
lll
V [ll1l
^
20
30
40
SO
60
70
80
90
Fig. 25.
100
no
120
130
140
ISO
160
28
A71 Introduction
to
the
Vanadium can be recognized in a mineral by a group of lines in the indigo at i22'3 to 3 24'5. Zirconium gives a brilliant spectrum of lines in the green, blue, silicate containing zirconium fused with lithium and violet.
carbonate shows the group of blue lines at 94-4 to 102, of violet lines at i39'4 and i40'7. Chromium is recognized by a group of three lines in i32'3, i33'7, and i35'6, and by the close triplet of lines at 7 4"3 to 74'5 ; with a single prism this will appear
single line.
and the
pair
the violet at
in the green
as a
strong
Carbon can be recognized in the spectrum of fused alkaline carbonates by the line at i34'5, which with increase of spark strength becomes a nebulous band. Without the Leyden jar this and the red line are the only lines seen the spectrum is best obtained by the use of a Leyden jar and air-break. The same spectrum is obtained from the spark between electrodes of graphite. The spectrum of selenium is shown in Fig. 25. It is obtained by taking the spark with Leyden jar between platinum wires covered with melted selenium, or between fragments of any of the selenides. The most characteristic lines are those in the green 7o'2, 7i'7, 72-4, 73'6, 75'8, 77'4> 79'5 and the three lines in the red: 35-1, 35-4,
; :
.;
36'4-
The most characteristic lines of tellurium are 36-3 in the red, 47*6 and 48-7 in the orange, 54 and 55-5 in the yellowish green, 57-5 and 59'5 in the green ; it is necessary to employ the jar. Sulphur gives both a line spectrum and a spectrum of bands, which are shown in Fig. 26. If the spark, with jar, is made to pass between wires of platinum covered with sulphur, or between fragments of sulphides, the most characteristic lines are the red group of five lines at 36 and 39, the green lines at 64-5 and 65-1, 68-5 and 696. The band spectrum of sulphur is obtained by passing the current from the induction coil through an exhausted tube containing sulphur which is heated by means of a Bunsen burner. In Fig. 26 are shown the spectra of the minerals berzelianite and argyrodite, a comparison of these spectra with the spectra of the various elements given, shows that berzelianite contains selenium, copper, silver, and thallium and argyrodite contains sulphur, silver, and germanium. It was in this mineral, one of the silver ores found at Freiberg, that germanium was found by Winckler in 1886. In the case of the elements which are gaseous at the ordinary temperature, or which are easily vaporized, the spectra are often very conveniently obtained by the use of exhausted tubes, known as Geissler or Thicker tubes. Such a tube is shown in Fig. 28. The electrodes consist of platinum wires melted into the glass, one at each end of the tube, and the central portion of the tube al> is made narrow. This portion of the tube-the " capillary "is placed in front of the slit, while the discharge from an induction coil passes through the tube. The colour of the discharge varies with the nature of the gas enclosed, e.g. with hydrogen the colour is red, with nitrogen violet.
;
spectra produced by
Means of
Electricity.
29
Sulphur.
i
i i i i
ii
hn
|i
fi!|i
iN|
ii
i[
iii
|
ill
pi!pLM
70
|
|
20
30
40
SO
60
80
90
100
110
120
130
140
ISO
160
[iin
|
iiii
|
iiii
|
iin|l^ifi|flfMfif^ifll1iljl1if
[
li^ptiflif^
90
100
110
20
30
40
50
60
70
80
120
130
140
150
160
iiiimii
iiiiiiiii
iiiTiV'iiiiiiLi niy^iHl'iffififliiLiiilimiiJ^
i
CofAper
^S.lM,
Silver
'*-
I
60
11
,7p,||tlO
II
I
90
|
100
no
120
130
140
ISO
160
III
II
iiii
[
Nil
nil
It
II
III
ml
II
\'^rpyroa.ic& Arpyrodite
1
n|j
|
mmiii
|
iiii|iiii[liii
iiTi
friT[iiTi
|
iiii' ^^
|
SP
\
60
II
70
I
II
,.80
90
|
I
100
i
I
110
120
130
140
150
160
Sulphur
II]
Mil
Bydrogert. vacu-um.
i!ii
|
iin|iiii
|
iiii
j |
nn|iiii|iiii
|
il!i
|
iiiimii|iiii
|
iiii[ iiii|iiii
[ini iin|
|
iiii[
iiii|iui[n!i ni
|
20
30
40
50
60
70
80
90
100
110
120
130
140
ISO
160
Helium.
|inr|
iiii
|
ni[|i!imiTMn
ii|iii
20
30
40
50
60
70
60
90
100
110
120
130
140
ISO
!60
ArffOiv. (red,)
iii1
|
!i[|
|
ln
i|
ii
ti |UrTTi|Ti^lili|lii
i [
i| ii
20
.
30
u-f<
40
50
60
70
80
90
100
110
120
130
140
ISO
160
--'<
-i'
30
71
Introduction
to
the
jii|iiii
|M^(ilf|fni[it^p!i|rtii|iii^iiiifrn^ii|frn|^^niii^
20
30
40
50
60
70
80
90
100
irO
120
130
1*0
150
r60
|i ii
ii
|
n Ni
|
i|i
iii if |tiyilr|
i
iii i|iiii| 1
iiLi|ilMi/ii|nii|nii
|
iiiifl^^
20
30
40
50
60
70
80
90
100
110
120
130
140
150
160
|i
in|
iiii
|
ill[
ii
|l'
lljfl
ll
L]1
'll
f|
li
1iii|iiimiii|iiii
[
ii
|
ii^
20
30
40
SO
60
70
80
90
100
110
120
130
140
150
160
|H
li
[N
ll
|l il l
|
ll
f
|
^|l
1f[
||
f|
||||
p||^|||||^||||f|[|
Li^
90
100
110
20
30
40
50
60
70
80
120
130
[40
150
160
pP"|"l|"ip1|i
20
ll|lNl|i"ipifiin[inl[inlLiiif[iii|^^^^
30
40
SO
60
70
80
90
100
110
120
130
140
150
160
111:
inlll]i
| |
BrcmaTte.
iiii
| |
Ill|niljl[ll|ll!mi ll|iiii[i1ii|iiii
iii1'|
y|rT
70
|llil| llir
iiii
[
ini|ini
|
iiii
|
'i[ii
iiiimii|i[ii
[
20
30
40
SO
60
80
90
100
110
120
130
140
150
160
nr ii| n
iji
ii i|
i|T ii l]
ifijiin]TiiMil|fi'i1|i11l l'iil|lii|^^
|
WiJ
60
70
Iodine.
20
30
40
SO
80
90
100
110
120
130
140
150
160
AhsorpiioTl
1llliini|INI|llll|IIITTTTTTTrTT]1 m]Triijiflt|iiii|ifli|iiii|Tni|i
PotassiiiTTV
Permxutganate.
iiip|iiiimii|iiii|ini i||ii|iii
ni|iiH|iiii|iiii|iiii
|
20
30
40
50
...
60
70
80
^
90
100
110
r20
130
lio
150
160
Didymium'
^^
jmpa|inip|iiij| ilip|
i
ii
i|ii
H
|
iii
|
nl^
20
30
.
40
50
60
70
I
80
90
.
I00
MQ
t
120
130 140
150
160
SrbuiTn Chloride.
20
30
40
50
60
70
80
90
Fig. 27.
100
irO
120
130
140
llo
160
spectra produced by
Means of
Electricity.
31
The pressure of the gas left in the it is yellow. tube is usually only a few millimetres of mercury ; if the pressure is too great the discharge refuses to pass, or takes place only as a thin thread-like spark, but as the pressure is reduced the light given out by the tube becomes brighter, and the discharge fills the whole tube. The colour is often different in the wider portions of the tube from that of the capillary, and the colour (and spectrum) of the negative pole, B, is often different from that of the positive pole. A, and different again from that of the capillary, ab. It is, however, the capillary portion always which is adjusted in front of the slit. The spectrum given by a Geissler tube is not always that of the gas present in the tube in the It is extremely difficult, in filllargest proportion. ing the tube, to avoid the presence of impurities, and it may happen that the electric discharge picks out one of these impurities, and renders it luminous, so that the tube may give the spectrum of something quite different from the gas supposed to be contained in the tube. Thus it has happened more than once that the spectrum of some unsuspected impurity has been described as that of a new element, Tubes which or as a new spectrum of an element. have been exhausted by itieans of a mercury pump commonly show the spectrum of mercury, often to the exclusion of all else, and at high exhaustions the commonest spectrum is that of an oxide of carbon, due to dust in the tube, or to the grease used in lubricating the taps, so that the evidence obtained by the use of tubes only is to be received with very
and with helium
great cautioji. If the gas in the tube is hydrogen, the spectrum consists mainly of three bright lines, shown in Fig. 26. Hydrogen vacuum tubes generally show, besides these three lines, a large number of feeble lines filling up the whole spectrum ; these constitute the so-called secondary spectrum of hydrogen, but it is not certain
that they are
due
to
hydrogen
itself
It is
very
diffi-
Fig. 28.
cult to obtain absolutely pure hydrogen, and especially tube filled with water vapour to free it from all traces of water. gives practically the same spectrum as has been described. Helium is one of the gases discovered by means of spectrum The most characteristic line in its spectrum is an extremely analysis. yellow line, a little more refrangible than the sodium Ime.
brilliant
in the This hue was first seen in 1868, at the time of a solar eclipse, no but Sun, the of chromosphere and prominences the light from known. substance of terrestrial origin giving the same line was must be an Nevertheless, it was concluded by Lockyer that there
32
An
element characterized by this particular line, and he gave to it the name " helium," from the Greek word for " Sun," and for twentyseven years helium was known only as an element existing on the sun. But in 189s, Sir W. Ramsay obtained from the mineral clevite a new gas, which turned out to be the long-sought helium. Since then this new gas has been obtained from other minerals, and has been found to exist in small proportion (one in a million) in air. At a pressure of 7 or 8 millimetres the helium tube glows with a brilliant If the tube yellow light, and gives the spectrum shown in Fig. 26. is more highly exhausted, the green line at 83 '3 becomes much the brightest, and the tube emits a brilliant green light, but only if the When mixed with other gases, helium is best detected gas is pure. by the green line in tubes highly exhausted. Argon is another gas existing in the air, discovered in 1894 by Lord Rayleigh and Sir W. Ramsay ; it appears to constitute about I per cent, of the atmosphere. A vacuum tube containing argon glows with a red light the spectrum is complex showing chiefly red lines and only feebler lines in the blue but if the tube is connected
;
with a
Leyden
jar,
the
spectrum changes, the red lines having become fainter, and many new green and blue lines having appeared. The behaviour of oxygen in a vacuum tube is peculiar it is not to be detected when mixed with other gases, and as already remarked, air in a vacuum tube shows the same speqfrum as nitrogen. A tube filled with pure oxygen, when not much exhausted, shows a continuous spectrum with a thin thread-like spark. As the exhaustion is increased, four lines make their appearance from the capillary, and these get brighter as the tube is further exhausted, while the glow from the negative pole spreads into the capillary, so that we now have the four lines superposed upon the special spectrum of the negative pole. If a Leyden jar and air-break be used, the four lines disappear and give place to the " elementary line spectrum," which is mapped in Fig. 16, No. 9. The four lines constitute the " compound line spectrum " of oxygen; it is mapped in Fig. 26. It appears that oxygen is capable of giving no less than seven
;
different spectra.
Nitrogen in a vacuum tube gives a spectrum of bands, shown in No. T. It shows a number of bands of one character in the orange and yellow, and an extensive series of bands of a different character in the blue and violet. This spectrum is entirely different from that given by the gas at atmospheric pressure when the spark with Leyden jar is sent through it. The line spectrum then obtained forms part of the air spectrum shown in Fig. 16. There is also a different band spectrum seen at the negative pole, so that it appears that nitrogen, like oxygen, is capable of giving several different
Fig. 27,
spectra.
an- three
Besides argon and helium, Ramsay and Travers have found in other new gases, neon, krypton, and xenon, the spectra of which are shown in Fig. 27. Of neon there appear to be one or two
spectra produced by
Means of
Electricity
33
parts in 100,000 parts of air; the spectrum has a yellow line a little more refrangible than the characteristic helium line, and a mass of
left
of
this,
There is about as much krypton as helium in the air one in 1,000,000 parts. The most characteristic lines in the spectrum are a bright green line at 6o'2, and pairs of lines in the red, orange, blue,
and
violet.
Xenon
the blue.
is
occurs in
still
smaller quantity
its
Chlorine seems to give only one spectrum a line spectrum. This obtained either by the discharge through a vacuum tube containing chlorine, or by sparks in the gas at ordinary pressure, or from fused The vacuum tube gives chlorides, or minerals containing chlorine. The the best result with chlorine at a pressure of 50 to 100 mm. lines by which the presence of chlorine is best detected are four lines in the yellowish-green at 64 to 67, a bright green line at 74, which is seen to be double on employing a higher dispersion, and a group of blue lines at about 95, in which three lines are very strong. A vacuum tube containing bromine shows a spectrum of lines the same lines are seen with fused bromides or minerals containing bromine when the spark is made to play upon them. The brightest green lines are a group of blue lines at 94'4j Q^'I) ioi'i, and 102 8
;
and
76-4.
richer in lines than the spectra of principal lines are at 48'3 in the red, a group in the yellow from 53 to 58, lines in the green at 62'8, 62-9, 63"9, 65, 6'.'2, 68'6, 68'9, 76^5, besides many blue lines.
The spectrum
of iodine
is still
chlorine
and bromine.
The
CHAPTER
Some few
IV.
P'lG.
slit of the instrument ; and Fig. 30 represents a glass trough to contain the solution to be used instead of the test-tube of It is better, however, to use a wedge-shaped trough, since F'ig. 29. in this way the effect of different thicknesses of the absorbing medium on the spectrLmi is seen. Fig. 31 shows the absorption spectrum of a solution of chromium chloride, of which a dilute solution appears green to the eye, while a strong solution or a greater thickness of The red end of the spectrum is solution appears purple or red. shown to the right in the figure, and the position of the dark lines of the solar spectrum are marked below, B in the red, D in the yellow, E and /' in the green, and so on. The bottom of the figure corresponds to a small thickness, and the top to a much greater thickness. The
Absorption Spectra.
The
Electric Arc.
35
small thickness cuts out some light in the yellowish-green and in the blue, and the light transmitted is therefore of a green colour, but as
II
Clil'SEDB
Fig. 31
Fig. 30.
the thickness or strength of the solution increases, the absorption increases until at last nothing but red light makes its way through. In Fig. 32 we have the absorption spectrum of a solution of potassium permanganate ; a very weak solution
shows
iive
and
F, but as the thickness or strength of the solution increases, all the light in this region is absorbed and the bands disappear, a solution of moderate strength being purple,
and a very strong solution allowing nothing but red light to pass. This same spectrum, as shown by a weak solution, is drawn again in Fig. 27, No. 8; it must be understood that in such a drawing of an absorption spectrum the black marks represent dark Some colourless or all but colourbands. less solutions show well-marked absorption
bands, often useful in detecting
stance.
the sub-
Fig. 32.
Thus didymium and erbium salts spectra. are easily detected by means of their absorption shows the absorption spectrum of a solution of didymium
in dilute solution; the strongest
Fig. 27 chloride
at 73-9
bands are
at 53 to 55,
to 74'6.
the
quite position of these dark bands in didymium salts is not band being slightly in the sulphate as in the chloride, each the acetate the bands are shifted towards the red, and in a solution of remarks that "these Bunsen red. the still further shifted towards didymium compounds differences in the absorption spectra of different any general theory cannot in our present complete state of ignorance of media, be connected with other for the absorption of light in absorptive alterations in phenomena. They remind one of the slight gradual
...
and
The
same
36
An
is
Introduction to
tlic
undergo when the weighted, or of the change of tone which an organ pipe exhibits when the tube is lengthened." Bunsen further calls attention to the fact that the molecular weight of the acetate is greater than that of the
pitch which the notes from a vibrating elastic rod
rod
Fig. 33.
sulphate,
and
is
heavier than
that of the chloride. Fig. 27 shows the absorption spectrmii of most conspicuous bands are at 7 3 '4 and 91 'o.
Coloured
complexity.
gases
also
give
absorption
spectra,
of
great
Bromine, iodine vapour, nitrogen peroxide, iodine monochloride give innumerable dark lines. The relationship of
Absorption Spectra.
The
Electric Arc.
37
these absorption spectra to the emission spectra of the will be discussed later on.
same substances
Another mode of ignition of the substances to be examined is by the use of the electric arc. Unless we have at command the supply of electricity from a dynamo, we must employ batteries, and as at
Grove or Bunsen are necessary, the method somewhat inconvenient. The carbon, whether volatilized or not,
least forty cells of
is is
carried across from the positive pole to the negative, thus producing a "crater" in the positive pole. Fig. 33 shows the incandescent poles of the arc; they may be projected upon a screen by the
arrangement shown
in
Fig.
34.
To
obtain
steady arc
some
Fig. 34.
Fig. of regulating lamp must be used. 35 represents Foucault's electric lamp ; in this the carbons are moved by two trains of clockwork, the one of which causes the carbons to approach each When the arc is in other, and the other causes them to separate. exact adjustment, both trains of clockwork are held by a catch governed by the armature of an electro-magnet working against a spring. If the carbons come too close together, the electro-magnet
form
draws the armature forward and releases the set of wheels which the carbons separate from each other; if, on the other hand, the carbons get too far apart, the current is enfeebled and the spring draws the armature back and releases the other train of wheels, but so long as the distance between the two carbons is exactly right, both
make
38
A/i IntroductioH
to
the
Study of
Spectruiii Analysis.
As the positive carbon wastes away wheels are locked. about twice as fast as the negative carbon, the clockwork is so made An alternating as to move the former twice as fast as the latter. current may also be used to produce the arc; in this case both carbons wear away at the same rate. There are many other forms of automatic regulators for the arc, but some form or other is necessary.
trains of
l-'ic:
The spectrum of the electric arc is, in the first place, a continuous spectrum due to the incandescent solid carbon. Then, in the next place, it gives certain bands, the same as seen in the spectrum of the Bunsen flame (see Fig. 16, No. 9), which are believed to be due to the vapour of carbon, the temperature of the arc being that due to
Absorption Spectra.
39
the volatilization of carbon, or perhaps 3500 C. It must be observed that the temperature of the arc is much less than that produced by the spark from the induction coil, either with or without Leyden jar and as a result the arc-spectra of the metals are usually much simpler than their spark-spectra. Thus in the arc-spectrum there are no lines due to arsenic in the portion visible to the eye, whereas in the spark-spectrum there are some 50 lines. In the complete arc-spectnmi of zinc there are about 73 lines, but in the spark-spectrum
something
measured.""'
like
400
lines
have
been
In order to produce the arc-spectrum of a metal, the lower carbon is hollowed out to form a cup, in which fragments of the metal or of one of its salts are placed, we then obtain bright lines of the metallic vapour superposed upon the continuous spectrum of the carbon The upper pole is a rod of poles. carbon, or sometimes a rod of the metal conunder examination is used. venient arrangement for producing the arc- spectra of the metals is shown in Fig. 36, in which there are six hollow carbons, any one of which may be brought under the top carbon rod, and thus into the right position for obtain-
The continuous spectrum due to the carbon is fortunately much weakened by the introduction of metallic subvolatile
stances into the arc, more particularly if metals such as cadmium are
being employed. Then the presence of lines due to metals other than those intentionally introduced and resulting from impurities in the carbon rods employed must always
FiG. 36.
be expected
iron,
in particular lines
due
to
"
are * The arc- and spark-spectra of the elements (according to wave-lengths) work. A comparison w. 1 show eiven together in the tables at the end of this [e.g. barmm, arsenic, antony that in tome cases the spectra are very different are more numerou than chromium, etc.), and that while generally the spark-lmes arc-spectrum con ains lines not seen the arc-lines, there occur cases in which the is observed When extremely exact measurements are made there lines in the spark. he .1 compared with the same a very slight shift of the lines in the spark as This is attributed to greater pressure in the arc, a sli|ht increase of wave-length. spark than in the arc.
40
A//
Iiitrodiictioii
to
the
carbon bands already spoken of. Even if the carbons have been heated for hours in a current of chlorine, the lines of iron and calcium are still to be seen. In Fig. 21, which shows a portion of the arcspectrum of strontium, the first line on the left is due to calcium, the next two are due to aluminium, and the next to calcium. These four " impurity " lines can be recognized in all arc-spectra.
CHAPTER
LENGTHS.
V.
THE DIFFRACTION SPECTRUM. MEASUREMENT OF WAVEwe have only spoken of one method of producing a spectrum, by the use of prisms. But there is another method and a very important method of producing a spectrum, by the use of what are
So
far
viz.
In order to understand this it will be necessary to consider at some length the nature of light, and the mechanism by which it is propagated from a luminous source to the
eye.
The velocity of light has been measured by several different methods, which give concordant results. Romer, in 1676, by observing the eclipses of Jupiter's moons, found that light required 16 minutes and 26 seconds to traverse the diameter of the earth's orbit, or 185,600,000 miles. Bradley determined the velocity of light from the aberration of The velocity of the earth in its orbit combined with the velocity of light has the result of causing a star (apparently) to describe a small circle or ellipse once in a year round its true position, the diameter of this circle is foundtos ubtend an angle of 20'445 seconds of arc, and the velocity of light is found by dividing the velocity of the earth in its orbit by the tangent of this angle. Fizeau, by an arrangement of a rapidly revolving toothed wheel,
light.
found
how
long
it
and
to
his
Foucault, employing a rapidly revolving mirror, found it possible measure the velocity of light in a room through a distance of only m n Foucault's apparatus is shown in Fig. 37. a few metres.
represents a rapidly revolving plane mirror ; the source of light is at S, and an achromatic lens, K, is so placed as to throw an image of M', of which the centre is at 0, S somewhere on the circle R' S' the axis of the revolving mirror. If this image, after reflection at the revolving mirror, is formed at /, and if a concave spherical mirror is
s\ the light is sent back placed at M, whose radius of curvature is again to the revolving mirror and reflected back again to S, if the mirror has not had time to move sensibly while the light was passing from to s' and back again, a b'vs,&. plane, unsilvered plate of glass, Observation shows that there is a and O the lens of a micrometer.
42
An
Intrcductioit to the
Study of
Spectriuii Analysis.
measurable displacement of the image when the mirror is in rapid motion, the ray of hght which started towards R' entering the eye at O as if it came from S'. By placing a tube filled with water, T, between the revolving
mirror and a concave mirror at M', it is possible to measure the velocity in water, and it is thus shown that in water the velocity of light is just three-quarters of the velocity in air. Foucault's method has been employed with improved apparatus by Michelson. His result is 186,330 miles per second, or 299,890,000 metres per second, or 2 '9989 X 10'' kilometres per second. The result obtained by Cornu was 3*004 by Fizeau, y^S by Romer, 3'io; and from the aberration of light, 3"o83.'"' Light consists in a very rapid vibration of a highly elastic medium, called the luminiferous ether, which fills all space, and pervades all transparent substances. Light travels from a star (or the sun) to the earth by means of waves, . * in which the particles
;
<
represented in Fig. tn' 38, the motion of the par' * ' tides in the row a a' is n Fig. 38. communicated to the next row, which move down, and come back to their positions of rest at c c\ then move up and come back to their original positions at lib'. The length ab is called a wave-length, and the time of oscillation of any one particle is the time required for light to travel from a to /'.
**,.'[
This
is
Listing, from a discussion of all Cornu's measurements, thinks the most probable value to be 2-999 and Newcomb has obtained 2-998. The value may therefore be taken as 300,000 kilometres per second.
:
The
Diffraction Spectrum.
Measurement of Wave-lengths.
light falls
43
when
it
upon a
undergoes a change in direction, part being reflected according to the law of reflection, and part passing into the glass, being refracted according to the law of refraction. The explanation of these phenomena given by the undulatory theory of hght is as follows : when the wave of light reaches the
surface of the reflecting and refracting surface, each point of that surface becomes the centre of a disturbance, which spreads out in each medium, but with less velocity in the glass than in the air. Let in Fig. 39 represent the plane surface of the glass, and sa, s'b the
MN
Fic;. 39.
direction of an incident beam of light, of which the wave-front is at When the wave-front reaches the surface at b, a right angles to sa. spherical wave starts from /' as centre, and by the time the wavefront has got to a, the new wave has a radius bd equal to ca in the original medium, and in the glass the new wave has a radius of only two-thirds of ca, since the velocity of hght in glass is only two-thirds
of that in air. If now from a in the plane of incidence we draw tangents to the two spherical surfaces at d and/, these will be the wave-fronts of the reflected and refracted beams respectively ; the direction of the reflected beam will be bd, and that of the refracted
beam
It
bf.
medium, the triangles abc equal to the angle bac, that is, the incident and reflected rays make equal angles with the reflecting surface MN, and, therefore, also with the normal, aV ; and the
will
easily
be seen
abd
is
44
An
equal to baf,
ab'
is
,.
is
ab
:
sine of the angle of incidence, therefore, bears to the sine of the angle of refraction the constant ratio ac bf, which is the law of refraction. The ratio ac bf is called the index of refraction, usually denoted by it is simply the ratio of the velocity of Hght in the first
The
fj.,
medium
second medium. have now to consider the phenomenon known as the interference of light. When amy portion of the luminiferous ether is affected by two waves of equal intensity, they may combine so as to produce increased motion and, therefore, greater intensity of light, or they may neutralize each other. If in Fig. 40 two waves come
to the velocity in the
'"
We
Fig. 40.
together so that they are in the same phase, that is, so that the crests and troughs of the one wave coincide with those of the other, we but if the one wave is behind the shall have an increased effect other by half a wave-length, or by an odd number of half wave-lengths as shown in Fig. 41, the two disturbances will neutralize each other
;
Fig. 41.
and darkness will be the result. This result may be reahzed experimentally by the arrangement of Fresnel, shown in Fig. 42. A beam of sunlight enters a dark room by a narrow slit in the shutter, and, after passing through a plate of red glass, falls upon a cylindrical lens, L, of short focus, and is thereby reduced to a narrow line of red
light at S.
at
The light from S falls upon two plane mirrors Im lm\ which meet a very obtuse angle, and after reflection proceed as if it had come
s
s'
The line in which the the line of light at S. The spherical waves which proceed from s and s' as centres are in a condition to
from the two points
mirrors meet
interfere.
* There is no light in directions oblique to the reflected and refracted rays, IjGcause disturl^ances in these directions destroy each other by mutual iirterference.
is
parallel
Measurement of Wave-lengtiis.
45
circular arcs represent successive positions of these wavefronts, each arc being further from the origin by half a wave-length
than the preceding arc. Upon a screen placed at right anglesto the line Aa we obtain a series of fringes alternately dark and light. There will be increased light wherever a dotted arc crosses apoUier dotted arc, or where one drawn full crosses another; but there will
full.
It will
be
Fig. 42.
seen that where the screen meets the line k.a we shall have the central bright fringe, and on each side of this we shall have a dark fringe at the points determined by the condition that the distances from and / differ by half a wave-length for red light, i.e. at the points n ?i; we have also dark fringes at the points ', whose distances from s and J differ by three-halves of a wave-length, but at the points
j-
cc, whose distances from s s' differ by one or two wave-lengths, i.e. by an even number of half wave-lengths, we have bright fringes. If, therefore, we can measure the distances ns ns' we can determine the length of a wave of red light. This is found to be less than the thousandth part of a millimetre, or denoting the one-thousandth part of one millimetre by the symbol ix, the length of a wave of red light is found to be o"656 /a, and of violet light about o"4 /x. A different method of producing these diffraction fringes is by the use of a " grating." Such diffraction gratings were first employed by Fraunhofer, who made them by winding very fine wire round two parallel screws of very fine thread and equal thickness, or by ruhng parallel lines on glass with a diamond point by means of a dividing
engine.
The
finest gratings
now
are those
known
as
Rowland
gratings,
46
yht Introduction
to
the
ruled by a fine dividing engine on polislied metal, having in one inch 14438 lines (or s68'44 to a millimetre). In Fig. 43 let s represent a luminous point, and / a screen and let the circular arc nXm represent any position of the spherical The effect at / is wave-front (whose origin is s) between j- and p. the same as if each point of this arc were itself the origin of a wave. From/ as centre let a series of circles be drawn with radii increasing by half a wave-length, the first having the radius /A, the next pa, then pb, pc, and so on. We thus divide the wave-front A; into " elements of interference." The element Aa is larger than ab, ab larger than he, and so on, but the elements approach more nearly to Now, considering the equality as they get further away from A.
Fig. 43-
Fig. 44.
effect at/, for each point in ah, there is a point in A(7 half a wavelength nearer/; interference will take place, and the effect produced by ab will be completely neutralized by A(7, but the effect produced by A(7 will be only partially neutralized by ab. In the same way be is partly neutralized by ed, and so on, and the elements of interference,
taken in pairs, will neutralize each other more completely the further A they are. The light which reaches P from s, therefore, comes altogether in the direction sV ; none comes in the direction mp, i.e. light travels in a straight line, and not round a corner. But the case is different if at ki/i a grating is interposed so as to stop off the alternate elements of interference, then light will reach P in the direction ;;/. This result is, perhaps, seen more clearly from Fig. 44, in
away from
which
SAC
at Aahe.
The
length, t'C
represents the direction in which light falls upon a grating distance dC being greater than eC by half a wavegreater than dC by the same amount, and so on, light
Measurement of Wave-lengtJis.
making an angle
is
47
dCK
SAC.
If cr
be drawn
dCk
f-
dr
and cd
is
the
Hence
is
equal to the
dc
there are
miUimetre, cd
=:
r.
and sine
d<Zh.
length in millimetres. To an eye placed at C, then, the appearance represented in Fig. 45 will be presented ; the line of light at S is seen
t
just as
it
miiin
Fig. 45.
it would be if the grating were not there, but on each side of be seen a dark space, followed by a spectrum having the violet This is called the nearest the centre and the red furthest off. spectrum of the first order, and beyond it may be seen a spectrum of the second order, with the colours just twice as widely spaced as The spectrum of the third order overlaps that in the first spectrum. of the second, and the fourth, fifth, etc., overlap the others more and more, so that their colours become more and more indistinct. The measurement of the exact wave-length of the lines of a spectrum is a problem of the highest importance, and one to which To a vast amount of care has been devoted by many observers. give some idea of the accuracy with which these measurements can now be made, the following measurements of the wave-lengths of the Fraunhofer, using two wire sodium yellow lines may be quoted gratings, found 5888 and 5896, while with a glass grating he obtained 5888 ten-millionths of a millimetre, for the mean of the two lines. Angstrom, of Upsala, in 1868 published his " Recherches sur le Spectre Solaire," in which he arrives at the values s89S'o and 5889-0 for the two sodium lines. Van der Willigen Ditscheiner obtained 5905-3 and 5898-9
will
scale of wave-lengths, in which the chemical origin of a large number of the Fraunhofer lines was marked, and this map and the
on the
it
many
years as the
This " spectre normal " of Angstrom marked a very great accuracy ; but the construction of more accurate gratings in advance by Rowland, and the publication of a photographic map made by
Rowland with
needed
his
new
gratings,
showed
for the
that
revision.
Rowland's values
48
An
Other measuvements of the less refrangible sodium line are 5896-25 by Miiller and Kempf, 5895-90 by Kurlbaum, 5896-20 by The most probable value is 5896-161 Pierce, and 5896-20 by Bell. in air at 20 C. and 760 mm. or in vacuo 5897-8. The distribution of the colours in the diffraction spectrum is the blue different from that in the spectrum produced by any prism end is more compressed in the diffraction spectrum, and the red end This is clearly shown in the proportionally lengthened out. is No two prisms can be relied upon to give coloured frontispiece.
; ;
exactly the
same
distribution of colours,
and
it
is
therefore neces-
sary to reduce the readings obtained with a prism-spectroscope to the standard scale of wave-lengths ; we "must therefore explain how this may be done. The following table shows the relative
positions occupied by the Fraunhofer lines B, D, E, F, and G, in dispersion spectra produced by prisms of 60 of crown-glass, flintglass, and carbon disulphide, with which are compared the positions
of the
same
lines in a
between B and
is
in
spectrum obtained by diftVaction. The interval each case divided into 1000 equal parts.
'truin.
Meas
Scale
number.
Lithium
31-8
.
Sodium
50'o
.
Thallium
67-8
74-5 105-5
Magnesium
Strontium
so
An
on the curve the wave-lengths of the Fraunhofer Hnes, and so determine their positions approximately on the scale of the spectroscope.
directing the instrument to the sun or to a bright cloud, the Fraunhofer lines will certainly be found at or near these positions. Now let these Fraunhofer lines be read off as exactly as possible, and from their positions, and those of the lines of the flame-spectra, let a more accurate interpolation-curve be drawn, and let this curve be used to find the positions of the lines of the spark-spectra. The final curve should be drawn when the positions of these spark-lines have been carefully observed. If it is not convenient to make use of the spark-spectra, a very fair curve may be constructed from the lines of the flame-spectra and from the Fraunhofer lines, but a little trouble in obtaining as accurate a curve as possible will be well repaid. As a sample of what may be done with a one-prism spectroscope and reflected scale, the following numbers, taken from Lecoq de Boisbaudran, for the wave-lengths of bismuth lines, are compared with
Thale'n's
Lecoq de
On
numbers
Lloisljaudran.
Tliale'n.
Lecoq de Loisbaudran.
Thale'n.
5 143 '5
5123-5
The lines from which Lecoq de Boisbaudran's interpolation-curve was drawn are the following
:
Scale reading.
Wave-length.
Scale reading.
Wave-length.
Potassium Solar a
.
Solar
Lithium
Hydrogen
Cadmium
Zinc
.
Lithium
Sodium Copper
b5'55 72-50 77-81 80-78 S3-7I 86-25 SS-oo 94-15 100-00 103-25 105-90 109-00
Silver Silver
Thallium
.
Cadmium
Hydrogen
Cadmium
Strontium Iron
.
j>
Lead
Fig. data.
5700 5607
114-00 118-40 124-40 > 30-03 141-75 152-83 157-60 174-28 180-So 188-25 200-83 216-33 220-75
5464 5349 5208 5085 4861 4677 4607 4383 4307 4226 4101 3968 3933
46
is
methods of measuring the positions of the lines of conveniently be put into two groups, which may be called methods of consecutive coincidences and methods of simultaneous coincidences. The chief plans employed are the following
different
The
a spectrum
may
Consectttivc coincidences.
(i)
-(2)
(3)
The graduated arc and vernier. The tangent-screw micrometer. The bright-line micrometer.
Measurement of Wave-lengths.
Simultaneous coincidences.
(4) (5) (6)
(7)
The reflected scale. The double-wire micrometer eyepiece. The divided lens micrometer. The photographic method.
lO
so
00
toe
lij
2-Jj
1^0
Fig. 46.
Some
and others
of these methods are more suitable for a small spectroscope, for a large one; or, again, particular methods may be
52
An
employed in one case and not in another. For exanaple, cross-wires can be employed with the solar spectrum, or with any spectrum
of
sufficient
brightness,
while
they
are
useless
spectra.
favourite plan with the opticians is that of the divided arc and vernier ; in which the telescope carries cross-wires, the intersection of
is brought first to coincidence with one line, then with a second, and so on. This, of course, is a method of consecutive coincidences, and it is a necessary condition of obtaining correct results that the coUimator and slit shall remain rigidly in the same position, and that the cross-wires of the telescope and the vernier shall retain the same relative position during the motion from one These conditions are attended to in the massive line to another. construction adopted by Steinheil and some other continental makers, but are fatally disregarded when the instrument is constructed of slender metal, and when the collimator and observing telescope, instead of being firmly grasped at the centre of gravity, are merely
which
screwed by one end into a slender upright of brass, further weakened at the most important point by being attenuated into some (so-called) ornamental shape. Certain precautions must be observed in the use of a spectroscope The eyepiece should first with cross-wires to obtain good results. be removed and so adjusted that, on looking through it at a sheet of white paper, the cross-wires are seen in sharp focus ; then, replacing the eyepiece in the observing telescope removed from the spectroscope, the telescope should be exactly focussed on some distant
object.
mator should then be adjusted till some lines in the green say b in the solar spectrum are in accurate focus. The instrument is then in adjustment." When used on the red or blue portion of the spectrum, the focus may be adjusted with the observing telescope, but the collimator should not be altered. It is necessary that the ray to be measured should be in exact
focus together with the cross-wires. If this is not the case, the ray will alter its position slightly with reference to the cross-wires, if the eyes be slightly moved. The adjustment may therefore be tested by moving the eye slightly and observing whether the ray and the crosswires move together. There is also a slight movement of the rays consequent on lateral shifting of the source of light ; this is less the narrower the slit is, and the more distant the source of
light
is.
instruments are provided with a tangent-screw micrometer a long screw, the head of which is divided into a hundred equal parts, by means of which a slow motion can be given to the observing telescope, and the number of turns of the screw, and parts of a turn necessary to carry the cross-wires from one line to another, is noted.
Some
that
is,
"
For a
dift'eient
Measurement of Wave-lengths.
;
53
In the bright-line micrometer * the image of a fine slit in a brass is seen by reflexion at the first surface of the prism, and so is superposed upon the spectrum the plate and slit have a slow motion given by a micrometer screw. This form of micrometer is specially useful with very faint spectra, when cross-wires would be useless. In observing with cross-wires a luminous spectrum, the lines of which are faint, it is necessary to admit a certain amount of light into the observing telescope, sufificient to illuminate the wires (conveniently by raising an edge of the cloth used to cover up the prisms). This general light renders very faint lines invisible. In all these methods of consecutive coincidences it is necessary that no shifting of the parts of the instrument by bending or shaking, nor any disturbance of the position of the source of light, nor of the exact position of the eye, should take place during the passage of the cross-wires from one line to the next. In the methods of " simultaneous coincidences " all these sources of error are avoided by observing at the same instant two lines one a known line, used as a reference line, and the other the line to be measured. The method of the reflected photographed scale, already described at some length, may be employed as a method of simultaneous coincidences, and so made more exact if, when the reading of any line is noted, care be taken to observe that the sodium line is still exactly at 50; or if the sodium line is not in the field, then that some other line used as reference line is exactly in its right position at the
plate
instrument for use with large spectroThis is an eyepiece the bifilar micrometer eyepiece. similar to those employed for astronomical purposes, provided with two crosses of fine spider lines in the focus of the eyepiece, which must therefore be of the Ramsden construction. One of these crosswires remains fixed ; the other is moved by means of a micrometer screw. The interval between the line to be measured and a line of
scopes
In thus be determined with great precision. taking an observation, a slight motion is given to the fixed cross-wires by means of the slow motion or tangent-screw of the observing telescope, the micrometer screw of the eyepiece being at the same time adjusted by the other hand, till the observer is satisfied that each line is truly coincident with the inter-section of the corresponding spider
lines.
Another device for measuring the interval between two lines, quite equal in accuracy to the bifilar micrometer, is that of the In this'instrument the micrometer screw divided-lens micrometer.f moves one-half of a lens placed just in front of the prisms, and The effect is to cause one-half divided along a horizontal diameter. of the spectrum to move along under the other half, and the sodium hne or any other convenient line is used as a substitute for the crosswires,
and
is
lines to
be
* Microscopical Journal, January, 1870. Proc. Physical Society, vol. t Phil- Mag., August, 1S75.
p. l6o.
54
^i" l72trodiichon
to
the
measured.
light
It will be seen that the necessity of admitting extraneous illuminate cross-wires is avoided, and this instrument can therefore be used with faint spectra with precision. The photographic method is, of course, a method of simultaneous coincidences, inasmuch as tlie positions of the known lines employed as references are recorded at the same instant as those of the
to
the divided-lens micrometer may have fitted to it a device for mapping the spectrum at the same time as the positions of For this purpose the steel rod on which the the lines are measured. micrometer screw is cut is made about three times as long as usual, and the extra length has cut on it a much coarser thread. On this there travels a little brass piece carrying a steel point, with which a We thus obtain a trace can be made on a slip of blackened glass. mark on the blackened glass corresponding to each line of the spectrum. Another instrument, very useful in measuring photographed spectra or in drawing maps of spectra from measurements, is This consists of a brass cylinder, on which Beckley's spectrograph. the photograph is stretched, and the edge of the cylinder is graduated and provided with a vernier. There is also a straight edge which can be brought down upon the photograph parallel to the lines of the Each line in succession is brought up to the straight spectrum. edge, and the position of the cylinder is read off by means of the vernier. The accuracy of reading is increased by substituting for the straight edge a small microscope with a three-inch objective, and with cross-wires in the eyepiece. Whatever the method of measurement, the numbers obtained must be reduced to the uniform scale of wave-lengths, so as to be independent of the peculiar construction of the instrument, the number, position, and refracting angle of the prisms, the dispersive power of the material of which they are made, of variations in the It is clear that in temperature, and of all other disturbing causes. such a method each line can be mapped only by means of its colour, that is to say, by the length of the wave of light by which it is pro-
duced and a spectrum so represented must be such a one as is produced by diffraction, and not by dispersion. Dispersion spectra obtained by the use of prisms of different materials vary greatly in the relative breadth of the colours, so that in mapping a spectrum it is by no means sufficient to give the positions of only two or three
;
Many
from
Three spectroscopes (each with a single prism and reflected by Duboscq and intended to be exactly alike, diflered as shown in the following table. The numbers show the difficulty of constructing two instruments with exactly similar
scales
:
Measurement of Wave-lengths.
5 5
^
56
At! lnti-oductio7i
to the
Then
is
K
We
XoWo)(Aj
(3
,,
j)(X,
Xj) X,)
(Xi]
X3:.)(X]
X.,)
(o
;/,)(Xi
X',)
""
;/,)
+ X,i T^i;
X)(;;
X,;?o
and
c={\The
;/,)
spectrum by Professor Hartmann will show the degree of accuracy which may be obtained. The constants were found to be
X
2232-8
ii
464-251
c=
658860
The
first
the second
column of the table gives the scale reading of the line, column its true wave-length, and the third column the
2232-8 -f
_6s88^o
464-251
_-,
Measurement of Wave-lengths.
57
advantage of enabling us to detect at once any reading inconsistent with the rest. If the positions of as many reference Hnes as possible are mapped on the paper, and a smooth curve is drawn through these points, or through as many as possible, and having the rest as near the curve as possible, and as many above the curve as below it, then one observation is corrected by another. The curve so drawn is more likely to give correct results than an
irregular
line
made up
off
many
straight portions so as to
pass
through all the points. In drawing the curve it will be found convenient to use a steel rule, which may be bent by the hands into the curve required this, however, requires the co-operation of two persons one to hold the rule down on the paper (stretched on a drawing-board), and the other to rule the curve with a finely pointed hard pencil. If oscillation frequencies be employed instead of wave-lengths, the line obtained is less curved, as may be seen from Fig. 46. This is an advantage, since the less curved the line is, the more easily is it drawn, and the more exactly can it be employed. A map of oscillation frequencies is intermediate between a diffraction spectrum and a dispersion spectrum, the red end being less extended when compared with the blue end than in a diffraction spectrum, and more extended than in the dispersion spectrum. A map drawn to wave-lengths is too much distorted to be advantageously employed with a dispersion spectroscope; and, on the other hand, a spectrum mapped with a dispersion spectroscope does not sufficiently resemble the same spectrum seen with a diffraction spectroscope but a map of oscillation frequencies, being intermediate between the two, is not so different from either but that it is suitable for use Further, a map of oscillation with a spectroscope of either kind.
;
frequencies exhibits most clearly relationships between the lines of a specturum ; as, for example, in the spectra of silver or copper, where pairs of hnes are repeated always with the same difference of oscillation frequency, or in the spectra of zinc and cadmium, where the same triplets are repeated again and again.
CHAPTER
VI.
ON THE PRODUCTION OF DARK LINES BY ABSORPTION. THE FRAUNHOFER LINES OF THE SOLAR SPECTRUM.
bright lines
a discontinuous spectrum of an incandescent gas, but that an Thus from the incandescent solid gives a continuous spectrum. light of glowing lime in the oxyhydrogen flame we get a bright continuous spectrum, in which the colours follow without break in succession from red through every shade of orange, yellow, green, and blue, into deep violet. If, now, the flame of a spirit-lamp coloured be with common salt which itself gives the bright sodium line brought between the glowing lime and the slit of the spectroscope, we The cause of see a dark line in the place of the bright sodium line.
The
rule has
been
laid
down
due
that
is
in general
to
Fig. 47.
this
phenomenon
is
which
it is
itself
in various ways.
able If a
sodium be exhausted and sealed up, then, on heating the sodium so as to fill the tube with sodium vapour, and holding it in front of a sodium flame, and if a spectroscope be used, and the the tube appears dark tube be held between the flame and the slit of the instrument, a dark line is seen exactly in the same place where the bright sodium
glass tube containing a few pieces of metallic
;
appears the moment the tube is removed. Or, by the use of the electric light, the dark line may be projected on a screen.
line
On
the Production oj
Dark Lines
by Absorption.
59
The observation of the exact coincidence of the dark line in the Sun's spectrum with the bright line of sodium seems to have been first made by Fraunhofer. So long ago as i8r4, Fraunhofer concluded that the dark lines of the Sun's spectrum were somehow produced in the Sun itself. He also observed that the light of the moon and the brighter planets (being only reflected sunlight) contained the same dark lines as seen in the direct light of the Sun. Foucault, in 1849, seems to have been the first to observe the dark sodium hne in the spectrum of the electric arc, but he did not explain how it was produced; and it is to Kirchhofif that we owe the clear explanation of the phenomenon, and its application to the discovery, by spectrum analysis, of the composition of the solar
atmosphere.
Fig. 48.
Kirchhofif mapped the spectrum of the Sun with very great care, using four prisms (Fig. 48), and giving the positions of the lines on an A comparison of these lines with the arbitrary scale of equal parts. bright Mnes of the elements showed that the following elements must be present in the sun sodium, calcium, barium, magnesium, iron, chromium, nickel, copper, zinc, and hydrogen. This list has been increased by subsequent researches by the addition of the following titanium, manganese, cobalt, carbon, vanadium, zirelements
:
:
conium, cerium, scandium, neodidymium, lanthanum, yttrium, niobium, molybdenum, palladium, sijicori, strontium, aluminium, cadmium, rhodium, erbium, silver, beryllium, germanium, tin, lead,
and potassium, and possibly also iridium, osmium, platinum, tantalum, thorium, tungsten, uranium, and oxygen.'gallium,
It may l)e * Fig. 50 is a copy of part of Kirchhoff's map, showing the /' lines. coinpaied with the photograph of the same portion of the spectrum shown in Fig. 47 shows a convenient arrangement in observing the absorption Fig. 51. spectra of gases.
6o
An
Introduction
to
the
The chief Hnes of the visible portion of the solar spectrum, as seen under small dispersion, from the red hydrogen line marked C is the double to the lines in the violet, are shown in Fig. 49 sodium line, E is a double line due to calcium and iron, ^ is a group
6\
mapping of the lines of the solar spectrum by Angstrom, Fievez, Cornu, and others, but all this has been superseded by the
photographs obtained by the use of Rowland's concave gratings. The complete map of the solar spectrum made by Rowland is something like forty feet long, and contains something like 20,000
lines.
One
spectrum
Fig. 5 2
;
of the
is
portion of the
is
reproduced
in
with this
may be compared
same
nj
/7t
1^3
<7e
fra
iliijl
yfi
yjKf
/e7
/(
/au
/m
^tj
/ks
mk. yMMi^i^AiH
T
Ft.
\
Fig. 50.
portion of the spectrum made by Mr. Higgs with a Rowland grajting, and reproduced by permission (Fig. 53). In Fig. 51 we have a reproduction of Mr. Higgs's photograph of the b group, in the upper portion of the plate, and below it a portion of the spectrum from the violet. The upper portion was photographed in the spectrum of the third order, and the lower portion in the fourth order. The scale of wave-lengths shown applies to the upper spectrum ; the wave-length of a line in the lower spectrum is exactly three-quarters of that shown by the scale. It is a valuable property of the concave grating that it allows the photographing of two portions of the spectrum at the same time, and on the same plate, the wave-lengths of which have an exact numerical ratio. Of the lines which can be distinguished in the upper spectrum, thirty-seven are due to iron, eleven to titanium, seven to nickel, and seven to chromium, four to magnesium, two to cobalt, and one each to calcium, vanadium, and yttrium. The b line most to the right (known as /',) is due to magnesium, 4 is also due to magnesium, b-. is a double
\
62
An
(N -te-
co
p
CD
iii^-
Tlie
63
''*''***^^^^''***^^^
r
-^
~m
^^r.
64
line
'
An
due
and
b^,
also a double,
is
is
due
to
magnesium and
iron. n.
65
66
An
Introduction
to
the
Wave-length.
5165- 416*
/J
5166- 454 5167- 497s 678s 5167- 885 5168- 123 5168- 360 5168- 32
45169- 069s
\5i69' 220s 5169' 469 5i69- 664 5i69' 871 5'7o- 271 5170' 655 5170- 767 5170- 937 5171- 192 5171- 778s 5172'3S6
/'5i72'856s
5173 499
5173'652
5173 917s
5174'077 5174'203
5174 595
5175'099 5175'423 5175'575
5176 735 S176 954 5177 179 5177 410 5177 577 5177 784 5177 979 5178 156 5178 644 517S 970 5179 293 5179 695 5179 958 51S0 233 SiSo 572 5180 747 5181 041
5181
334
67
Wave-length.
68
An
symmetry in the structure of the band is very interesting; it must indicate some connection between the vibrations of the molecules of oxygen which give rise to the absorption lines, and it is most
|[||J|jjj![[[i
ll|i
Fig.
5.
same structure reproduced almost exactly in the other group A, which lies further in the red. Fig. 57 is a reproduction (by permission) from a photograph of this group by Mr. INlcClean.* group of lines due to water-vapour in the neighbourhood of the lines, has been called the " rain-band" ; it is darker when there is much water-vapour in the air, and may be used to foretell rain with
interesting to find the
some
success.
It
shows Angstrom's map of the telluric lines of the solar spectrum; and Fig. 59 shows the rain-band. It is possible to distinguish the telluric lines from the real solar lines in the spectrum of the light which reaches us from the sun by an application of Doppler's principle carried out by Cornu. In consequence of the sun's rotation, the position of a true solar line is slightly different, according as the light which reaches the spectroscope comes from the advancing or receding limb of the sun. If a small image of the sun, produced by a lens, is thrown vipon the slit of the spectroscope, and if the lens is made to oscillate through a distance about equal to the diameter of the image, so that the slit is illuminated, first by the light from one edge of the sun, and then by that from
Fig. 58
* Olszewski found that a layer of liquid oxygen, 30 millimetres thick, gave an absorption band agreeing in position and character with the Fraunhofcr
rise to
Professor Jansseu, in 18S8, observing at Meudon in Fig. 57, the electric light at the top of the Eiffel Tower in Paris, observed the and li groups just as they appear in the "High Sun" spectrum. In travelling from Paris to Meudon the light would have to pass through about as much oxygen as the sun's rays would in traversing the atmosphere. In the invisible infra-red region of the solar spectrum, bands due to the carbonic acid and aqueous vapour of our atmosphere have been observed.
band A, shown
6g
llOU^.l l'm-.^.LLL
l
"^
JO
An
W**i
The
Spectm
71
the opposite edge, the solar hnes will appear to tremble in consequence of the slight alternately opposite displacements ; but if the lines are produced by absorption within our own atmosphere, it will
matter nothing from what part of the sun the light comes originally, and the position of the line will not be affected by the vibration of
the lens.
(T
J ft,.
J
Fig. 59.
Fraunhofer, in 1814, examined the spectra given by the brightest of the fixed stars, and recognized that they differed from the spectrum of the sun. Fig. 60 is a copy of a drawing of the spectra of two bright stars made by Huggins and Miller in 1864, giving the result of a direct comparison of the bright lines of some metals with the spectrum of the star. It thus appears that Aldebaran contains hydrogen, sodium, magnesium, calcium, iron, bismuth, tellurium, antimony, and mercury ; and that Betelgeux (a Orionis) contains sodium, magnesium, calcium, iron, and bismuth.
Spectrum.
The portion of the spectrum visible to the eye is only a very small portion of that which can be observed by the use of photography at the blue end, and by the use of the thermopile, or Bolometer, for
the red end. The portion of the solar spectrum which can be seen by the eye is little more than that shown in Fig. 49, extending from wave-length 6500 to 3900. By the use of photography lines of wave-length only 1000 have been observed, and by the use of the bolometer radiations These radiations of wave-length of 100,000 have been detected. beyond the red form the infra-red spectrum ; those beyond the visible Photographing with glass violet form the ultra-violet spectrum.
is
which would
is cut short at about 3176, with quartz be employed, the spectrum can be photographed as far as 2099. If ^ Rowland concave grating is employed, the image is formed by the concave mirror of speculum metal on which the lines are ruled, and all lenses are avoided ; but
With
flint-glass
the spectrum
if
72
Ah
Introduction
to
tlie
The
Fi-aunliofe7-
73
74
All Introduction
to
tlic
even then a limit is put by the absorption of the air, and of the gelatine used in making the photographic plates, as a support for the
sensitive silver salts.
The extreme limit of wave-length looo mentioned above was attained by Schumann by working in a vacuum, and with sensitive plates of emulsion on glass without gelatine. Beyond the red end of the visible spectrum the lines have been photographed, by the use of proper plates, as far as wave-length 12,000, wiih the diffraction grating, but beyond this point the heating effect of the spectrum must be utilized. Here prisms and lenses
must consist of rock-salt or other material transparent to dark heat. Professor Langley, in 1884, employing a large Rowland concave grating of short focus with a rock-salt prism and a bolometer, measured the heating effect in the spectrum down to wave-length
53.oo-
The Bolometer
lines of the
consists of a
fall;
spectrum
this strip
and consists of some metal whose resistance varies rapidly with change of temperature. Such an instrument will detect a change of temperature of one-ten-thousandth of a degree Centigrade. Fig. 6 [ shows Langley's infra-red solar spectrum. The dark line oj at 14,000 is due to water-vapour, as are also at 12 at 18,300, 26,400, and a strong band from 50,000 to 110,000 with maxima at 71,000 and 81,000. A band Y at 46,000 is due to carbon dioxide. Langley, working on Mount Whitney at an elevation of 12,000 feet, has measured 700 dark lines in the infra-red invisible spectrum.
bridge,
CHAPTER
VII.
76
yii/
Introduction
to
the
different These different types of spectra are believed to indicate According to the stages in the evolution of stars from nebute. the nebular hypothesis of Laplace, the stars have been formed by
2
'J
<
Z
r.
-.
l#
The
spectra of well-defined nebute exhibit bright lines, as if the radiating bodies were altogether gaseous, the hydrogen lines being prominent.
and Nebuhe.
77
4)
Ot-1
78
An
Introduction
to
the
stars) have spectra Certain stars (often called the "Wolf-Rayet" and heUum Dein^ consisting chiefly of bright lines, those of hydrogen conspicuous.
i*^l-jj
^11-
tlia
m
These, then, are the newest stars, after which follow in order of The stars or age the stars of types O, I, II, III, IV, in order. nebulce which have the highest temperature have the simplest spectra.
and
Ncbulce.
79
As
a star cools its spectrum becomes more complicated ; thus the temperature of Sirius is supposed to be higher than that of Arcturus or of our sun. In Fig. 62 (see also Fig. 66, No. 3) we have the spectrum of a Lyrae showmg the complete series of the hydrogen lines, and in Fig.
63 we have photographs of stars of Secchi's types II, III, and IV, taken by Professor Hale, with the 40-inch Yerkes telescope, and reproduced by kind permission of the Editors of the Astrophysical Journal. Sir William and Lady Huggins have published a magnificent
8o
A7t Introduction
to
the
"Atlas of Representative Stellar Spectra" (London, 1899), from which Figs. 65 and 66 are reproduced by permission. They show some of the principal changes which are supposed to have taken
place in the gradual evolution of stars. description from the "Atlas" :
We
"
stars,
We
which we regard as in the least condensed condition (omitting Wolf-Rayet stars). These stars exhibit strong absorption lines of
helium. " In the early white stars the series of hydrogen lines continues
8i
and distinct, to the end of the series, while, on the contrary, as the solar stage begins to set in, the ultra-violet members of the hydrogen series become less distinct, and when the full stage
is
reached have practically disappeared. " In Bellatrix and in /3 Lyrte a very strong sub-characteristic
the strong lines of helium, so that these stars may as Helium stars. In Rigel this sub-characteristic pair of lines at about 3860 which belong to silicon, and any similar ones might be known as Helium-
presents itself in be distinguished includes a strong so that this star Silicium stars. " One glance
at the ultra-violet region of a Cygni shows as its chief characteristic, next to hydrogen, a pair of strong lines and a thin line on the more refrangible side, forming a triplet at about 3760 (due to Titanium), and the pair as in Rigel at 3860. This star may appropriately be known as a Titanium-Silicium star. " The gradual incoming and the advance of the characteristics of the solar stage can be well followed in the photographs, becoming successively stronger as we pass from Sirius to the fainter star of great advance is made when we Castor, and then to a Aquite. reach Procyon, and the progress is still more rapid through y Cygni to the complete solar stage as it presents itself in Capella, of which the spectrum is not distinguishable from that of the Sun."
Sir J. in recent times photographed. Herschel, in 1824, gave a careful drawing of this object as it was seen in his large reflecting telescope of twenty feet focal length and eighteen inches diameter. Splendid work in the photography of the nebula has been done by Sir Isaac Roberts and Dr. Common, and at the Lick Observatory by means of the Crossley reflector. Professor Keeler concludes that there must be something like j 00,000 nebute in the heavens. Huggins found that while the stellar and resolvable nebulas gave spectra resembling the spectra of the fixed stars, the planetary and
irresolvable nebula; gave spectra consisting of bright lines, thus showing that in these bodies we have masses of incandescent gas. He observed generally three bright lines one due to hydrogen and coincident with Fraunhofer's line F, one supposed to be due to
nitrogen,
and the
third
was not
identified.
82
Att Introduction
to
the
In Fig. 67 is shown a nebula, resembling a sickle in shape; Fig. 68 shows the spiral nebula in Canes Venatici ; Fig. 6g, the spiral nebula 604, and Fig. 70 the annular nebula in Lyra ; Fig. 7 r,
the great nebula of Orion, photographed by Sir Isaac Roberts with an exposure of three and a half hours. Sir William Huggins thus describes his discovery of the true nature of the nebute "I was fortunate in the early autumn of 1864 to begin some observations in a region hitherto unexplored, and which, to this day.
Fig. 67.
remains associated in my memory with the profound awe which I felt on looking for the first time at that which no eye of man had seen, and which even the scientific imagination could not foreshow. " The attempt seemed almost hopeless for not only are the nebulas very faintly luminous as Marius puts it, like a rush-light shining through a horn but their feeble shining cannot be increased in brightness, as can be that of the stars, neither to the eye, nor in
'
'
the spectroscope, by any optic tube, however great. " The view of the nebula; as parts of a fiery mist out of which the heavens had been slowly fashioned, began, a little before the middle of the century, at least in many minds, to give way before the revelations of the giant telescopes which '* come into use, and especially
spectra of
tite
Stars
and
Nebula'.
83
Rosse.
"Nebula
after
nebula
yielded,
being
Fig. 68. Spiral nebula in Canes Venatici. {From Morgans " AdvaJiced Physiography.")
Fig. 69.
innumerable stars, as the optical power was increased and so the opinion began to gain ground that all nebulfe may be capable of According to this view, nebute would have to resolution into stars.
;
84
All Introduction
to
the
be regarded, not as the early stages of an evolutional progress, but rather as stellar galaxies already formed, external to our system cosmical sand-heaps too remote to be separated into their component stars. Lord Rosse himself was careful to point out that it would be unsafe from his observations to conclude that all nebulosity is but the glare of stars too remote to be resolved by our instruments. "On the evening of August 29, 1864, I directed the telescope I looked into the for the first time to a planetary nebula in Draco. A single bright No spectrum such as I expected spectroscope. At first I suspected some displacement of the prism, and line only that I was looking at a reflection of the illuminated slit from one of
'
' ! !
Fig. 70.
Ring nebula in
"
Lyra.
[From Morgan's
Advanced Physiography.")
its faces. This thought was scarcely more than momentary ; then the true interpretation flashed upon me. The light of the nebula
so, unlike any other light I had as yet' subjected to prismatic examination, could not be extended out to form a complete spectrum. After passing through the two prisms, it remained concentrated into a single bright line, having a width corresponding to the width of the slit, and occupying in the instrument a position at that part of the spectrum to which its light belongs in refrangibility. A little closer looking showed two other bright lines on the side towards the blue, all the three lines being sepa rated by intervals relatively dark. " The riddle of the nebute was solved. The answer, which had come to us in the light itself, read Not an aggregation of stars, but a luminous gas. Stars after the order of our own sun, and of the
:
and
Nebulce.
85
stars, would give a different spectrum; the light of this nebula had clearly been emitted by a luminous gas. With an excess of caution, at the moment I did not venture to go further than to point out that we had here to do with bodies of an order quite different from that of the stars. Further observation soon convinced
Fig. 71.
Photographed by
"
Advanced Physiography,")
life
is
me
that,
far
too minute
relatively to cosmical events for us to expect to see in succession distinct steps of so august a process, the probability
is
any
indeed overwhelming in favour of an evolution in the past, and still going on, of the heavenly hosts. A time surely existed when the matter now
S6
/I
Introduction
to
the
condensed into the sun and planets filled the whole space occupied by the solar system, in the condition of gas, which then appeared as
a glowing nebula, after the order, it may be, of some now existing in the heavens. There remained no room for doubt that the nebulae, which our telescopes reveal to us, are the early stages of long processions of cosmical events, which correspond broadly to those required by the nebular hypothesis in one or other of its forms. " It is necessary to bear distinctly in mind that the old view which made the matter of the nebulae to consist of an original fiery
mist, in the
'
could no longer hold its place after Helmholtz had shown, in 1854, that such an original fiery condition of the nebulous stuff was quite unnecessary, since in the mutual gravitation of widely separated matter we have a store of potential energy sufficient to generate the high temperature of the sun and stars. "The solution of the primary riddle of the nebula: left pending
I-"iG.
72.
Planetary nebtila
in
Aquarius.
some secondary
questions. What chemical substances are represented by the newly found bright lines? Is solar matter common to the
nebula; as well as to the stars? What are the physical conditions of the nebulous matter? " Further observation showed two lines of hydrogen ; and recent observations have shown associated with it the new element recently discovered by Professor Ramsay, occluded in certain minerals, and of which a brilliant yellow line in the sun had long been looked upon as the badge of an element as yet unknown. The principal line of these nebulfe suggests probably another substance which has not yet
in
terrestrial
rocks by the
different
The
different
kinds
of nebulae,
no doubt, represent
planetary nebute show edges sharply defined in the form of a circle or slight ellipse, but a higher type still is that of the stellar nebulaj, in which
The
and
Nebulie.
87
In Fig. 72 is shown the planetary nebula in Aquarius, with its spectrum of three lines, and in Fig. 73 is shown the spectrum of the
88
/I
Fig. 74. Central pans of the great nebula in Andromeda. Photographed with the two-foot reflector of the Yerkes Observatory, by G. W. Ritchey.
(From
reflector of the Yerkes Observaan exposure of four hours; and in Fig. 75, a photograph taken with the Crossley reflector of the Lick Observatory, on July 6, 1899, with an exposure of three hours, of the remarkable trifid nebula in Sagittarius.
and
Nebidce.
89
The latest determinations of wave-lengths of nebular lines are given in Lick Observatory Bulletin, No. 19. They are
5006-89 4959'05 486 1 '5 Hj8, due to hydrogen. 4740-1 4685-8 447i"7 due to helium.
43634
4340'6 4101-9 4068-8 4026-7 3970-2
H7, due
HS,
to
hydrogen.
due to helium. He, due to hydrogen. due to helium. Hf, due to hydrogen.
39677
3965-1
|-2
H,
one nebula shows all of these lines. Spectroscopic evidence leads to the conclusion that the different chemical elements as we know them on the earth are widely diffused throughout the universe, and also no doubt others with which we have not yet made acquaintance upon the earth. The example of the gas helium whose existence as a new element was first inferred from the observation of the bright yellow line in the spectrum of the sun's chromosphere, but which has now been found upon the earth, teaches us caution in supposing that we have as yet made acquaintance with all the chemical elements. Only a very small proportion of the lines photographed in the spectrum of the sun have been identified.* No doubt some of the others are due to some of these as yet undiscovered elements. Chemists and physicists have been gradually coming to the conclusion that the elements are in some way connected together, as would be the case, for example, if they were all modifications of the same primary material. The name " protyle " has indeed been given to this hypothetical primary " stuff." The relationships between the properties of the elements exhibited in the statement of the " periodic law " make it almost certain that there is some such connection between the different elements. The relationships which have been shown to exist between the different lines of one and the same element, and between the similar lines of chemically related There is some evidence that elements, lead to a similar conclusion. certain lines observed in the spectrum of the star t Puppis form a series due to hydrogen, but these lines have not yet been obtained from
No
hydrogen
to
in the laboratory.
The Fraunhofer
seen
at
lines
and K, due
in solar
calcium,
are
frequently
the
highest
lines
levels
been determined.
90
All Introduction
to
the
prominences, accompanied only by the lines of the very light gases hydrogen and helium, and without the strong calcium line at 4226"9, while in the Bunsen flame this line is strong, but H and K do not appear. Schuster and Hemsalech have observed that in the electric discharge the elements of small atomic weight, such as magnesium and aluminium, are projected from the pole with greater velocity than heavier ones, such as zinc, cadmi\im, and mercury, and that in
Fig. 75.
Trifid
Nebula
Sagittarius.
{From
some cases
certain lines of an element are projected with greater velocities than other lines of the same element. These facts seem to show that the particles whose vibrations give rise to certain lines
in
the
particles
spectrum of an element may exist independently of the which give rise to other sets of vibrations.
Professor J- J. Thomson attributes the spectral lines to the vibrations of "corpuscles" much smaller than the chemxal atom, and he believes that the chemical elements are different aggregations
and NebuUe.
91
diffused
material. Possibly the original uniformly matter of the universe was protyle, and we have the evolution of the chemical elements going on and at different stages
in the different nebulae
same kind of
and
stars.
maintains that the differences between our terrestrial spectra and those of the sun and stars can only be explained by assuming that our elements are at stellar temperatures dissociated into finer forms of matter, and that "if the terrestrial elements exist at all in the sun's atmosphere, they are in process of formation in the cooler parts of it." This would account for the absence of some of our elements, if in the intense solar heat they are dissociated into simpler forms the spectra of which we do not
recognize.
Sir
Norman Lockyer
At the same time there are observations which lead to a different conclusion. Professor Rowland did not find any lines common to several elements, or where there appeared to be such coincidence
more accurate measurement showed some slight difference in wavelength, or revealed the presence of a common impurity. Sir William Huggins has investigated the spectrum of calcium and magnesium
his experiments seem to show that variations in density or in the relative abundance of material have an important influence on the relative behaviour of the different lines of a spectrum. Sir William says " In the laboratory, as well as in the stars, owe strongly marked feature which distinguishes some conditions of the spectrum of
:
the relative behaviour of a line in the blue at 4227, and in the solar spectrum as and K. As this relative behaviour is usually different in the spectrum of the arc as compared with that of the spark, the condition of relative feebleness of the blue line, which is characteristic in the latter case, has been regarded as a criterion of a higher temperature. As this view led to the
is
calcium
the lines
known
improbable conclusion that the temperature in the highest regions of the solar atmosphere, notwithstanding that the temperaturegradient must be extremely rapid near the solar surface, was very much greater than lower down, and just above the photosphere, we were induced to make in 1897 a series of experiments which showed clearly that the relative behaviour of the blue line to the lines and K, which had been attributed to an exalted temperature, could be brought about by reducing the density of the calcium vapour. " This view at once brought the interpretation of the solar phenomena into agreement with our knowledge, from other sources, of the probable physical conditions existing near the sun's limb.
"
We
by
this interpretation
of the relative
and K lines of the behaviour of the blue calcium line and the same substance, with an important criterion which we can apply to stellar spectra, since we meet with these lines of calcium in different conditions, as a marked feature in the gradual transition of stars from the white to the solar type. "What the spectroscope immediately reveals to us are the waves
92
Ah
Introduction
to
the
which are set up in the ether. As a rule it is only when a body is gaseous and sufficiently hot that the motions within its molecules can produce bright lines. The spectra of the heavenly bodies are to a great e.xtent absorption-spectra, but we have to study them through the corresponding emission-spectra of bodies brought into the gaseous form and rendered luminous by means of flames or of electric discharges. There does not appear to be any direct relation between the luminous radiation as shown in the spectroscope and the temperature of the flame, or of the gaseous contents of the vacuum-tube, that is, in the usual sense of the term as applied to the mean motion of all the molecules. In both cases, the vibratory motions within the molecules to which their luminosity is due are almost always much greater than would be produced by encounters of molecules having motions of translation no greater than the average motions which characterize the temperature of the gases as a whole. The temperature of a vacuum-tube through which an electric discharge is taking place may be low as shown thermometrically, but the vibrations of the luminous molecules must be violent, and it is to the fierce encounters of the few molecules which carry the discharge that the luminosity is due ; if all the molecules had similar motions the temperature of the gas would be very high. " So in flames where chemical changes are in progress, the vibratory motions of the molecules which are luminous may be very different from those corresponding to the mean temperature of the
is therefore called for when we attempt to reason by the aid of laboratory experiments to the temperature of the heavenly bodies from their radiation, especially on the reasonable assumption that in them the luminosity is not ordinarily associated with chemical changes, or with electric discharges, but is due to a simple glowing from the ultimate conversion into molecular motion of the gravitational energy of shrinkage. " The presence in the solar spectrum of lines which we can only jjroduce electrically is, however, an indication of the high temperature
of the sun."
Doppler's Principle.
If a radiating body, emitting vibrations of a certain definite frequency, is in rapid motion towards or from the observer, the frequency will be altered to the observer in consequence of the change in the velocity relative to the observer. Thus, a vessel moving in a direction at right angles to the waves will encounter more waves in a given time if she is moving against the waves than she will if moving in the same direction as the waves. The same thing is true of waves of sound and of light. When a railway engine, whistHng, passes an observer at full speed, he notices that the pitch of the whistle is higher when the engine is approaching, and lower
when
it is
If the velocity
spectra of
tlic
93
of sound in air is iioo feet per second, then an open organ-pipe of two feet in length will give a note of ^^, or 275, vibrations per second. If this note be sounded on an engine travelling 40 miles an hour, or 59 feet per second, then, when the engine is approaching.
the observer will hear a note of i^, or 289, vibrations per second, if the engine is receding the note will be -'^, or 260, vibrations per second ; which is an alteration of the note by about a semitone, up or down. If, then, the exact alteration of note can be measured,
and
known
velocity
94
All Introduction
to
the
of sound. The same reasoning applies to light, only here we are The velocity of light dealing with an enormously greater velocity. being known to be 300,400,000 metres, or 186,000 miles, per second, if the exact amount of shifting of the lines of a spectrum can be determined, then the velocity of approach or of recession between
The disthe source of light and the observer can be calculated. placement of known lines in the spectrum at the edge of the sun, or in sun-spots, indicates rapid motion of the radiating gas to or from the earth of 50 miles a second or more; in other words, we have here evidence of violent storms upon the sun. Huggins was the first to apply Doppler's principle to the measurement of the velocity of stars in the line of sight. His observations showed that Sirius and a Orionis are approaching us with a velocity of about 20 miles a second, and that Arcturus and a Lyrse are receding from the earth with a velocity of 40 or 50 miles a second. But more exact measurements have been obtained by the application During the last ten )'ears exact measurements of of photography. the velocities of stars in the line of sight have been made in the observatories at Potsdam and at Greenwich, as well as in America. At the Lick Observatory a special solidly built instrument the Mills
spectrograph, which is shown in Fig. 76 is employed for these measurements, with the great refractor of 36 inches. It consists of a collimator of i^ inch aperture and 28^ inches focal length, and a camera, rigidly mounted with their axes parallel. The light from the star after entering the slit is dispersed by a train of three glass prisms rigidly clamped for the minimum deviation of the hydrogen line F. A comparison spectrum from a hydrogen vacuumtube, or from an iron spark, is photographed upon the same plate, so that the resulting photograph shows the spectrum of the star lying between two comparison spectra, the lines of which serve as standards to determine the displacement of the lines.
CHAPTER
THE SUN,
The
VIII.
a central
body from which we get a continuous spectrum, and that surrounding this there is an absorbing atmosphere giving rise to the dark lines. By observations made during total eclipses of the sun, and in other ways, we become aware that outside the reversing layer there is an atmosphere of incandescent hydrogen called the chromosphere, and
Fig. 77.
beyond
this
again a
halo
known
as the cprona.
central portion of the sun which gives rise to the continuous spectrum sometimes called the photosphere need not necessarily be solid, for a sufficiently thick layer of gas under pressure would The produce a continuous spectrum like a solid or liquid does. heated gases must be in a state of violent motion, convection currents
The
g6
All Introduction
to the
carrying the substances up into the cooler portions of the atmosphere there to condense and fall back again into the hotter regions, where they are again vaporized. As a consequence of these violent motions the incandescent gas is projected into enormous " red
in full
dayhght by Professor
[From
the
I'lC. 79.
Eruptive prominence,
(From
flames," or prominences, which are best seen during a total solar eclipse, when the body of the sun is hidden by the moon. They vary form and in height, sometimes extending as much as 70,000 miles beyond the sun's disc. Fig. 77 gives a representation of the
prommences
The
Sun its
Prominences,
etc.
97
The Sun
its
Corona.
99
The Sun
its
Spots,
etc.
lOI
method of observing the prominences at any time, and not only during an eclipse, was discovered independently by Janssen and Lockyer, which consists in so weakening the white light of the sun by passing it through a long train of prisms that the monochromatic
Fig. S2.
visible,
is
now
possible
to
see
prominences round the sun at any time. of photographs of prominences made at the
and by the adoption of and to photograph the Figs. 78 and 79 are copies
Kenwood
Observatory,
Fig. 83.
[From Thornton' s
''
Advanced Physiography.")
Chicago, and Fig. 80 is a copy of a beautiful photograph taken at Fig. 81 shows a drawing of the corona at this the eclipse of 1900. same eclipse made from photographs (copied by permission from the strophyskal Journal) The telescope shows on the surface of the sun dark spots, varying
I02
A;/ Inlroduction
to
the
greatly in number and size, and often large enough to be seen with They share in the the naked eye, or with the aid of an opera-glass. general motion of the sun's surface, which is due to a rotation of the sun on its axis in about 27^ days, but they have also motions of their own ; they open out or close up and disappear, to be succeeded by
new
spots.
The
spots are
sun's equator,
between the
more abundant in two zones parallel to the latitudes 5" and 40, as shown in Fig. 82.
The bright surface of the sun is seen in a telescope not to be uniformly bright, but to present a mottled appearance. The brighter portions of irregular shape are called faculre. Fig. 83 shows the appearance of a portion of the sun, showing spots and faculpe. Fig. 84 shows the comparative size of the earth and of the great spot of
I"io.
84. The
great sun-spot of April 19, 1892 area, about 1,093, 500, 000 square niiles. (From Morgan's " Advanced Physiography.")
;
April, 1882.
A spot appeared in 1851, which was 140,000 miles in diameter. If, using a spectroscope attached to a telescope, the image of a spot is made to fall across the slit of the spectroscope, the Fraunhofer lines are seen to be darker in the spot than elsewhere, but they are also seen to be thickened and distorted, as
shown
easily
in Figs.
85 and 86.
Fio.
85. Thickening
be understood that these distortions are the result, in accordance with Doppler's principle, of rapid motions in the
It will
absorbing material
i.e.
of
an uprush or a downrush of
gas,
and from
the amount of disturbance of the lines the velocities of these motions can be calculated. The velocities measured amount to 50 or more miles per second and prominences have been seen to form with a velocity of 250 miles a second. When the slit of the spectroscope is placed on the edge of the
;
The Sun
its
Spots,
etc.
103
sun, the dark lines are seen to give place to the bright lines of the
I04
line
An
Introduction
to
the
is also frequently curved and displaced, showing rapid motion in the gas producing it. It is further to be remarked that all the lines
Se-i;
???^
.c
-~-
l--rt
-=
t;
iPf~-
I'""
I"-
C[uite quiescent, gives the bright lines of hydrogen and the bright yellow line D3 due to helium. The
same substance are not new Potsdam spectrograph. The chromosphere, when
of the
equally displaced.
Fig. 89
shows the
The Sun
chromosphere
its
Cromosp/iere, Corona,
etc.
105
is, however, disturbed by prominences, and by the slow formation of " domes," which may last for weeks, and may be due
Fig. 89.
{From
to the welling up of vapours from beneath. observing on Mount Sherman, has catalogued
io6
the
An
Introduction
to
the
chromosphere, due chiefly to calcium, iron, sodium, and magnesium. The great prominences seem to be formed over the spot-zones, the equatorial and polar regions being apparently regions of less activity. Respighi remarks that " there is a great difference in the Some develop and disappear in a few duration of prominences. They originate, minutes others remain visible for several days.
;
generally, in rectilinear jets, either vertical or oblique, very bright and well defined. These rise to a great height, often to a height of at least 80,000 miles, and occasionally to double that height ; then,
fall again upon the sun like the jets of a fountain. they spread into figures resembling gigantic trees, more or less rich in branches." Since 1869, when the spectroscope was first used for its investigation, there has been much discussion as to the nature of the corona. During the eclipses of 1869 and 1870, it was shown that the spectrum of the corona was discontinuous, and characterized by a In the eclipse of 187 bright green line 1474 on Kirchhofif's scale. it was conclusively shown by photography that the corona belongs to the sun, and that the appearance is not produced by our own atmoThe corona sphere, as had been maintained by some observers. was found to be not simply a soft effulgence of light, but to have a It exhibits rays, compared to those produced by certain structure. searchlights through a fog ; from the polar regions the rays appear sometimes curved, bending away from the axial line. The spectrum of the corona is found to be a faint continuous spectrum crossed by bright lines, the most noticeable being the line 1474 of wave-length 5317. A later measurement of this line in the corona by Lockyer gives the wave-length 5303; its origin is unknown, but it has been attributed to a hypothetical substance,
bending back,
Then
" Coronium."
belonging
to helium,
Other bright lines are 4230 and 4290, and others hydrogen, and calcium.
CHAPTER
NEW STARS. DOUBLE
IX.
STARS.
COMETS.
One of the most remarkable phenomena of the heavens is the occasional appearance of a new star. The earliest recorded instance is that of a star which appeared suddenly in the constellation Cassiopeia in 1572, and which was as bright as Jupiter. It increased in brilliancy so as to be visible even in the daytime ; its brilliancy gradually diminished for a year or more, and it has never since been seen. In 1866 a star of the ninth magnitude in the constellation Corona Borealis suddenly flashed up into a star of the second magnitude, and lasted for some weeks. Its spectrum was nearly this is recontinuous, but showed also bright hnes of hydrogen presented in Fig. 90.
;
Fig. 90.
Spectrum of
Corona
Borealis.
In 1876 a new star appeared in Cygnus which had never been seen before. It was of the third or fourth magnitude and of a deep Its spectrum was like that of t Corona Borealis, gold-red colour. but the continuous spectrum faded and the bright lines remained longest one line at 4990 is one of the lines characteristic of nebulfe
;
(Fig. 91).
In 1885 a new star appeared in the midst of the great nebula in It reached the seventh magnitude and then faded. Its spectrum showed In 1892 a new star appeared in Auriga. the bright Hnes of hydrogen and lines at 5750, 5010, 4960, and 4360, which have been observed in the spectra of some nebulae. Nova Sagittarii appeared in 1899 as a star of the eleventh Examined with the spectroscope, it showed a faint conmagnitude. tinuous spectrum crossed by nine bright lines, the same as seen in
Andromeda.
1900.
Its
spectrum was a
faint
io8
A// Introduction
to
the
continuous one with three bright bands agreeing in position with the three brightest Unes in the spectra of the nebulpe. On February 21, 1901, Dr. Anderson discovered a new star in
1876.
1877.
Fig. 91.
in
Cygnus.
the constellation Perseus of the second or third magnitude. On February 23 it had become brighter than Aldebaran ; it then declined in brightness, but increased at the end of March, and then
New
Stars.
Double
Stars.
Comets.
109
declined again till it became about equal to a star of the tenth or eleventh magnitude. Its colour at first white changed through claret-red to dull orange-red. At first its spectrum strongly resembled
H^, HC, K, He,
H/3
Feb.
Feb.
28.
Mar.
6.
Mar.
15.
Fk;. 92.
Spectrum
Ihc
of Nov.i Persei.
[From
that of
Astrophyskal Journal.)
Nova Aurigje, a continuous spectrum with dark lines and a spectrum of bright lines the bright-line spectrum increased in brilliancy, and the continuous spectrum faded. The principal bright
;
no
lines
A// Introduction
to
the
were of considerable width, and the hydrogen lines were The displacement of the dark lines indicated a especially briUiant. Velocities relative motion of something like 500 miles per second. which nearly as great have been observed in the solar explosions
give rise to the great prominences. Fig. 92 shows the spectrum as photographed on five occasions, from February 27, 1901, to March 28, 1901, at the Yerkes Observatory with the great telescope.
Fig.
93.
September
20, 1901.
very similar to the first spectrum of Nova lines are bright and broad, and accompanied on the more refrangible side by dark lines, as in Nova Auriga. If the appearance of a " Nova " is due to an outburst of hot gas, which cools as it recedes from the star, the cooler gases approaching us would give a spectrum of dark lines slightly shifted towards the blue, and the hotter gases on the side of the star remote from us would
Aurigc-e, the
The spectrum
hydrogen
Neiv Stars.
Double
Stars.
Comets.
This
is
II
give bright lines slightly shifted towards the red. what occurred.
exactly
may be
to have taken place in the spectrum gathered from a study of the photographs
K lines
no
is,
In the photograph of February 27, the and are bright, crossed by fine dark and lines. The band doubt, due to the superposition of the calcium and hydrogen
;4>
112
An
Introduction
to
the
This nebulosity is found to be expanding with the incredible something like ii minutes of arc per annum. These astonishing changes can be studied by a comparison of the two photographs, Figs. 93 and 94, taken on September 20, 1901, and February It has 8, 1902, with the two-foot reflector of the Yerkes Observatory. been suggested by Kapteyn that the changes seen are due to changes of illumination of stationary nebular matter. If this theory is correct, and if the successive illumination of portions of nebular matter further away from the centre of disturbance is to be explained as due to the time taken by the light of the central explosion to travel out to these outlying portions of matter, then we may draw some conclusions as to the distance of this Nova from the earth. Light travelling with a velocity of 186,000 miles per second, or 365 X 24 X 60 X 60 x 186,000 miles a year, we have to calculate at what distance from the earth this number of miles would subtend an angle of 1 1 minutes of arc. This wouldgive the distance of the Nova as about 1,833,1 60,000,000,000 miles. It would take light about 317 years to traverse this distance, so that the explosion which was seen on the earth in 1901 must have taken place in 1584, in the reign of Queen Elizabeth. On March 16, 1903, a new star was discovered in the constellation Gemini. It was of a red colour, and its spectrum showed bright lines on a faint continuous spectrum, the red hydrogen line being especially bright it corresponds pretty nearly with the spectra
velocity of
;
of
Nova
Auriga: and
Nova
Spectroscopic Binaries.
telescope enables us to see that some stars which appear as naked eye are really double. This may be either that the stars happen to be nearly in the same line of sight as seen from the earth, or because the two stars have some physical connection, the line joining them being perceived to change its position, indicating that the two stars form a system revolving round each other in a certain definite time. Sir "\\'illiam Herschel observed about 2400
The
one two
to the
and many more have since been added to the list. In the time of a complete revolution occupies many years, in other cases only a few days or hours. The star Mira (o Ceti) has a period of eleven months. For two-thirds of this time it is a faint star of the ninth magnitude ; it then becomes brighter, and is seen
double
stars,
some cases
second magnitude for ten or twelve days. Algol Persei) has a period of sixty-nine hours. Ordinarily of the second magnitude, it decreases gradually to the fourth magnitude, remains so for twenty minutes, and then gradually recovers its original
as a star of the
(/3
brightness. In many cases the second star of the double cannot be seen, and its presence is inferred from the variation in brightness.
The
variability of Algol
is
New
Stars.
Double
Stars.
Comets.
113
The introduction of the spectroscope has made possible the recognition as binary systems of many stars which could not be recognized as such even by the most powerful telescopes. The telescope can detect a system, like that of Algol, only in the exceptional case when the plane of the orbit which one component describes round the other passes nearly through the sun, so that one component seems to pass over the other, and so causes an eclipse. Unless the plane of the orbit is nearly at right angles to the line of sight, the motion to or from the observer at the spectroscope will give rise to a displacement of the lines of the spectrum. If only one of the two components is luminous, there will be a periodic displacement of the lines; if both are luminous the lines will be periodically doubled. When they are in conjunction the spectrum shows nothing unusual, but when one component is moving away and the other is approaching us, the lines will be displaced in opposite directions, and all lines strong enough to be seen in both spectra will appear to be doubled. Binary systems which have been discovered in this way by means of the spectroscope are known as "spectroscopic binaries." The bright star Capella (a AurigEc) was discovered in 1899 to be a spectroscopic binary ; the spectrum of the principal component is of the solar type, and that of the secondary component is of the Photographs taken in 1896 show that the principal Sirian type. component had a velocity towards the earth of 34 kilometres per second on August 31 ; 54 kilometres on September 16 ; 44 kilometres on October 5 ; and only 4 kilometres on November 1 2. The period seems to be 104 days. The pole star has been found to be a spectroscopic binary with a period of four days. Fig. 62 shows the spectra of the two components of /3 Cygni blue star, much the brighter of the two with a spectrum similar to that of a Lyra, and a feebler companion of a yellow colour.
Spectra of Comets.
One of the most interesting applications of spectrum analysis in astronomy is to the determination of the nature of comets. Many hundreds of these mysterious bodies have been observed, some periodic, returning time after time ; others are seen once and never
seen again. From successive observations of the position of a comet, If the orbit is a parabola, the comet is orbit can be determined. seen once when it circles round the sun, and never again ; but if the orbit is an ellipse, the comet may be expected to make its reappearance after the lapse of a definite number of years. Fig. 95 shows This comet was discovered in 1682 by the orbit of Halley's comet. He found its orbit to Halley, the Astronomer Royal of the time. be an ellipse, and its periodic time to be seventy-five or seventy-six years, and therefore predicted its return in 1758; it returned punctually in 1758 and 1833, and should reappear in 1910.
its
I
114
An
Introduction
to
the
the most remarkable comets of the last fifty years was The orbit as Donati's cometf which was seen in 1858. of this comet is shown in Fig. 96. It attained its fullest development and greatest brilliancy in October. The tail extended 40 across the sky, but besides the principal tail it developed two smaller, nearly
that
One of known
Fig. 95-
as shown in Fig. 97. It appears as if the tail of a comet were formed of matter driven off from the head by some repulsive force possibly electrical emanating from the sun. Unfortunately there were no spectroscopes waiting for the comet in 1858, and Donati's comet will not return for 2000 years.
straight ones,
New
Stars.
Double
Stars.
Comets.
6'6 years.
115
Biela's comet was discovered in 1772; and 1826, and found to have a period of
was observed in 1805 It was seen again in 1832, but at its return in 1846 it was found to have broken up into two portions, each with a tail and nucleus, and when next seen, in 1852, the two portions had become widely separated. Since then it has never been seen, and probably never will be seen again.
it
rapid changes which may take place in a comet are shown in and 99, which are from photographs taken on two successive nights of a comet which appeared unexpectedly in March, 1892,
The
Figs. 98
known
No less than thirty-six unexpected comets as Swift's comet. appeared in the ten years from March, 1890, to March, 1899. The first application of the spectroscope to the study of comets
11
A// Introduction
to the
Fig. 97.
Fig. 98.
Fig. 99
New
was made by
Stars.
Double
Stars.
Comets.
:
117
Sir William Huggins Sir William Huggins in 1866. thus describes his observation of the comet of 1868 " I had myself, in the case of three faint comets in 1866, in 1867, and in January, 1868, discovered that part of their light was peculiar to them, and that the hght of the last one consisted mainly of three Intense, therefore, was the expectancy with which I bright flutings. directed the telescope, with its attached spectroscope, to the much
brighter comet which appeared in June, 1868. "The comet's light was resolved into a spectrum of three bright bands or flutings, each alike falling off in brightness on the more On the evening of the 22nd, I measured the refrangible side.
positions in the spectrum of the brighter beginnings of the flutings on
Fig. 100.
I was not a little surprised the next morning to find the red side. that the three cometary flutings agreed in position with three similar Some time flutings in the brightest part of the spectrum of carbon. different from carbon of spectrum the down mapped had I before, In some of these sources, chiefly from different hydrocarbons. spectra the separate lines, of which the flutings are built up, are
more distinct than in others. The comet bands, as I had seen them on the previous evening, appeared to be identical m
individually
with character in this respect, as well as in position in the spectrum, current of the flutings as they appeared when I took the spark in a I immediately filled a small holder with this gas, olefiant-gas. be arranged an apparatus in such a manner that the gas could when a attached to the end of the telescope, and its spectrum,
Ii8
A?i Introduction
it,
to
the
Study of Spectrum
Aiuxlysis.
seen
side
" Fortunately, the evening was fine ; and on account of the exceptional interest of confronting for the first time the spectrum of an earthly gas with that of a comet's light, I invited Dr. Miller to come and make the crucial observation with me. The expectation which I had formed from my measures was fully confirmed. The comet's spectrum, when seen together with that from the gas, agreed
1.
Solar spectrum.
2.
Spectrum of carbon
In olive-oil.
3.
Spectrum of carbon
in olefiant-gas.
4.
II.
5.
1.
in
Hame.
as part of its was emitted by luminous vapour of carbon (Fig. loi). "This result was in harmony with the nature of the gas found occluded in meteorites. Odling had found carbonic oxide as well as hydrogen in his meteorite. Wright, experimenting with another type ot meteorite, found that carbon dioxide was chiefly given oft" Many meteorites contain a large percentage of hydrocarbons. From one of such sky-stones, a little later, I observed a spectrum similar to that of the comet. The three bands may be seen in the base of a candleits
it.
secret.
'subtle
light
New
Stars.
Double
Stars.
Comets.
119
"In 1881, for the first time since the spectroscope and also suitable photographic plates had been in the hands of astronomers, the coming of a bright comet made it possible to extend the examination of its light into the invisible region of the spectrum at the blue end. On June 22, I was able to obtain, with an exposure of one hour, a good photograph of the head of the comet. It was under a great tension of expectancy that the plate was developed, so that I might be able to look for the first time into a virgin region of nature, as yet unexplored by the eye of man. " The plate contained an extension and confirmation of my earlier observations by eye. There were the combined spectra of two kinds of light a faint, continuous spectrum, crossed by Fraunhofer lines, which showed it to be reflected solar light. Upon this was seen a second spectrum of the original light emitted by the comet itself. This spectrum consisted mainly of two groups of bright lines, characteristic of the spectra of certain compounds of carbon. It will be remembered that my earlier observations revealed the three principal flutings of carbon as the main feature of a comet's spectrum
in the visible region. The photograph brought a new fact to light. Liveing and Dewar had shown that one of these bands consisted of lines belonging to a nitrogen compound of carbon. We gained the new knowledge that nitrogen, as well as carbon and hydrogen, exists in comets. Now, nitrogen is present in the gas found occluded in some meteorites. At a later date. Dr. Flight showed that nitrogen formed as much as 17 per cent, of the occluded gas from the meteorite of Cranbourne, Australia." Comet 1898 I (Perrine) was observed by Professor Wright to have a spectrum of the usual type, the three characteristic bands superposed upon a relatively strong continuous spectrum. Comet 1898 VII (Coddington) showed the same spectrum. Comet 1898 X (Brooks) showed the three bands, the one in the green being much brighter than usual, with a weak continuous spectrum. Comet 1899 a (Swift) had a spectrum resembling that of Brook's comet of 1898; it was photographed, and gave the following results
:
4883 bright line. 472 blue band, agrees with ihe fourth carbon band.
Iti;
(""l^' band,
,,
,,
,,
fifth
carbon band
jl'^g'
4313
421
faint
,,
,,
,,
,, ,, ,,
band,
,,
,,
413 faint band, 4100 4074 bright. 4052 bright. 4042 bright.
4oi9|bright. ^ 4014/
,,
,,
4i28'i.
3987 faint. 3879 bright, agrees with cyanogen 3883'S. 3869 veiy bright, agrees with cyanogen 387 1'S-
CHAPTER X
THE CONCAVE GRATING.
The
is
PHOTOGRAPHY OF
THE
SPECTRUM.
great advaaice in accuracy in recent measurements of spectra undoubtedly due to Rowland's invention of the concave grating. In order to attain accuracy Professor Rowland had first to devise a method of cutting a sufficiently accurate screw. The method he adopted for this purpose will be found described in the " Encyclopaedia Britannica" in the article " Screw." In order to produce an accurate screw, the screw, having been cut, is ground by means of a grindingnut as long as the screw, the two halves of which can be tightened together as required. It is excessively difficult to avoid all periodic error in the screw, and that due to error in the centring and graduation of the divided head. The dividing engine constructed by Professor Rowland is capable, when maintained at a constant
temperature, of making a scale of six inches in length, with errors at no point exceeding one hundred-thousandth of an inch. To make a grating of that length with 14,000 lines to the inch requires four days and nights of continuous labour, and the result is seldom perfect, on account, possibly, of wear of the machine or change of temperature. The grating is ruled on a concave spherical surface of polished speculum metal. The lines are the intersections of a series of parallel planes, one of which passes through the centre of the sphere. The relationship between the number w of lines (and spaces) in a grating and the wave-length \ of the light by which a diffracted image is formed in a spectrum of the order N, is given by the formula
NX =
where
/
oj
(sin
sin
ff)
the incident and refracted light with the normal to the surface of the grating. This formula applies to either plane gratings or curved ones. (The lower negative sign is to be taken if the incident and diffracted rays are on opposite sides of the normal.) If the radius of curvature of the grating be
6 are the angles
and
made by
taken as a diameter on which a circle is described, then it may be shown that a luminous point anywhere on this circle will have its image somewhere on the same circle. Rowland adopts the mounting
shown
in Fig. 102.
SE and SG
wooden beams on
the Spectrum.
12X.-
is a tubular rails fixed at right angles to each other. iron girder equal in length to twice the radius of curvature of the and E, which run on the grating. It is supported by carriages at In all positions, then, S will be on the circle of rails GS and ES.
GE
which
The is the diameter. grating being at G, and the slit at S, and the camera or eye-piece at E, the image of the slit will always be in exact focus at E. Since passes through the centre of cur-
GE
GE
- o, and becomes
0)
sin
i.
just
mean
a resolving
The be seen as separate, to their difference in wave-length. difference wave-length of the two sodium lines is 5893, and their To separate them a spectroscope must have of wave-length is 5-97. power of 988. The most powerful gratings have a
If a grating has resolving power of between 100,000 and 200,000. the Nth order, of spectrum the employ altogether n lines, and we thus attain the then the resolving power is proportional to N. with a snialler order higher of a spectrum a using same result by inch are number of lines. Gratings of 15,000 or 20,000 lines to the a larger number of lines having in advantage little is there but used, for Fig. 54 shows Mr. Higgs's apparatus to the inch than this. Figs. 51, 53, 55, photographing with a Rowland concave grating. photographs produced. and 56 show (imperfectly) the quality of the than the other. Most gratings give a brighter spectrum at one side consequence of some peculiarity in the shape
We
concentrated in the spectrum of of the grooves, most of the light is The rail SE may be other. the or side one on order particular one so that the wave-lengths, representing graduated in equal divisions which it is desired camera may be set to any particular wave-length The radius of curvature of the grating is usually to photograph. plates are 19 inches long and 2 inches 2 li feet, and the photographic
'"
" We put in the sensitive plate and Professor Rowland says Having exposed that part photograph. wish to we part the move to more, -^e have no once expose and position we move to another
:
It
may happen
that, in
122
^in Introduction
to
the
thought for the focus, for that remains perfect, but simply refer to the table giving the proper exposure for that portion of the spectrum, and so have a perfect plate. Thus we can photograph the whole spectrum on one plate in a few minutes from the F line to the extreme violet, in several strips, each twenty inches long, and we may photograph in the red rays by prolonged exposure. Thus the work of days with any other apparatus becomes the work of hours with this. Furthermore, each plate is to scale, an inch on any one of the strips repreThe scales of the senting exactly so much difference of wave-length. different orders are exactly proportional to the order. Of course,
the superposition of the spectra gives the relative wave-length. One important property of the concave grating is its astigmatism, i.e. that
Fig. 103.
the image of a point of light at the slit is a short line. Thus a small spark gives a spectrum of sensible width there are no longitudinal lines due to dust on the slit, since these are brought to a different
;
focus."
103 represents a spectrometer and plane grating arranged the measurement of wave-lengths. This method involves the accurate measurement of the angular deviation of the ray observed, and the determination of the exact value of the grating space. The grating space is never exactly the same throughout the whole ruled
for
Fig.
surface. Regular or periodic variations give rise to " ghosts," or false lines, first described by Quincke
the so-called
and
carefully
investigated by Pierce, and to differences in focus of the spectra on opposite sides, as investigated by Cornu ; or a portion of the grating
Photography
of the Spectrum.
itself.
123
If
may be
this
abnormal ruling be confined to a few hundred lines, the effect will only be to diffuse a certain proportion of the light without producing false lines or sensibly injuring the definition, but an error will be introduced into the estimate of the grating space if this be determined by counting the number of lines and dividing this into the total length of the grating. The abnormal ruling is more likely to occur at the commencement of the ruled space, and if it can be detected it should be covered up.
Professor Hartley, in Thorpe's " Dictionary of Applied Chemistry," gives the following account of his method of photographing the
spectrum " In photographing spectra ranging from 4000 to 2000 tenthmetres the most suitable prisms are of quartz, as first used by M. Cornu ; they are cut perpendicular to the axis, one of right-handed, the other of left-handed rotation, each of an angle of thirty degrees. They may be cemented together, back to back, by a drop of glycerin,
:
Fig. 104.
{From Thorpes
^'
or the one may be fitted to the collimator close to the lens, and the other in a similar manner in front of the telescope or camera lens. The lenses are of similar thickness, and of right- and left-handed
Instead of a telescope, a photographic camera of someis employed, the plate carrier of which is so inclined to the direction of the mean ray of the spectrum that it makes an angle therewith of nineteen degrees to twenty-four degrees (Fig. 104). By such means all the rays, inclusive of the wave-lengths given above, are focussed with accuracy on one plate, and a magnifying power of twenty-five diameters may be employed with advantage in examining the photographs. The camera back is constructed so that the normal position is angular, but it is also made to swing on a vertical pivot situated at its centre, the centre of the pivot being in the same vertical plane as the centre and front of the surface of the The collimator should be fitted with a graduated focussing screen. A connection between this and draw-tube and rack and pinion. the front half of the camera is made of leather, folded bellows fashion. This admits of a certain degree of 'side-swing,' with to-and-fro motion for focussing. The camera should be mounted on an iron
crystals.
124
--in
Introduction
falls
to
the
frame, which
into position
matter what be the length of the collimator, it should be supported for at least one-half of its length, and admit of being fixed at the minimum angle of deviation for the mean ray. An opening is cut in the collimator-tube of about one-third of its circumference in width, and two inches in length, which may be This enables the experimenter to judge closed by a sliding cover. in which direction to move the electrodes, so that the rays of the spark may fall exactly upon the slit and pass along the axis of the
mator
is
fixed.
No
coUimator. This is done by means of a card, which fits just inside the opening, and has a vertical line ruled at a point coinciding with When the spark is properly adjusted, a the centre of the tube. The spindle-shaped bundle of rays strikes the centre of the card. to 7775^ of a" i"ch in width, and it is well to protect slit should be The source of illumination, if it from dust by a plate of quartz.
\F}vm Tlwrte's
Cheim^Uy.
Fig. 107.
Clioiulstry.
Dictionary of Applied
[From
of Applied Chemistry.")
absorption spectra are to be studied, should be condensed sparks from metals aft'ording lines of known wave-lengths, and so numerous that they practically serve the purpose of a continuous spectrum. " Details of construction of the camera will be understood from Figs. 105 and 106, which show the instrument in plan and elevation, while Fig. 107 shows the back of the camera with the dark slide in position, and the shutter in front slightly raised. is a rigid shallow box, in front of which is the lens in its fitting N. L, which is a continuation of M, is triangular in form, and is movable in two directions. It moves to and fro by means of the screw K, and makes a greater or lesser angle with the mean ray by turning on the pivot situated at X. The screws A A serve to clamp the swing-back in position. The camera back is constructed to carry plates capable of taking several photographs, and the dark slide is moved up or down by the thumb screws B B acting on a pinion which works the rack on the
XN
'
'
The
front
Co7icavi
Grating.
125
edge of the dark slide DD. CC is the movable frame for the dark slide. The interior edge of the dark slide is shown in shadow by GG, while is the slit in the camera back through which the spectrum is projected. E and F represent the door at Ihe back and the sliding front respectively. In the elevation (Fig. 106) is seen the body of the camera, and door to the same, through which is
J,
seen the slit by which the spectrum passes to the sensitive plate. In the act of focussing a glass plate covered with some fluorescent material is placed in the dark slide, and the lines of the spectrum are
viewed by
reflection.
care is required to secure a proper adjustment of the instrument and obtain good photographs. The battery power, size of coil, and condensing surface must all be proportional to the
"Some
distance between the electrodes, in order that the rapidity of the passage of the sparks should render the light almost continuous. The coil should give a six-inch spark in air, and it is better to excite it with three storage cells than with five cells of Groves battery. The electromotive force of the cells should be i"8 volts. The current required to work such a coil is from 7|- to 8 ampferes ; three storage cells yield 15^ amp'feres, but with a resistance of \ ohm in circuit, the current is conveniently reduced. Electric light carbons have been used for the resistance. The smallest condenser to be of any use has 72 square inches of tin-foil on each side of a glass plate, and a margin of 3 inches. It is fixed in a wooden frame, with binding screws attached to the wood. Special methods have been adopted for examining the emission spectra of Fig. 108. solutions, and for observing absorption spectra. {From Thorpe's Two materials have been commonly used for " Dictionary of solutions, namely, electrodes of graphite and gold. Applied Chemistry.") Wicks of gold wire are made to pass just above the surface of the solution, which is contained either in a porcelain crucible or in a small U-tube, one limb being longer than the other
(Fig. 108).
" Graphite electrodes for the same purpose are cut to the shape of The wedges wedges, about \ inch in length and \ inch in width. are attached to platinum wires, and are fixed in glass tubes as shown in the figure the lower wedge dips into the solution, which is carried Capillary attraction to its upper edge by deep grooves or scratches. keeps the lower electrode moist at its upper surface. Generally speaking, there are only two lines of any other element than carbon visible in the spectrum of good graphite, and these are the first and third of the quadruple group in the magnesium spectrum, wave-lengths 2794"4 and 28oi'i. The second and fourth are very faintly seen,
;
Impure graphite
is
apt to
show
126
Introductio7i to the
In order to secure fine definition, a condensing lens in front of It should be of quartz, of 3 inches diameter slit is necessary. and 3 inches focal length. The spark is best if very brilliant and that is It should pass upwards not exceeding pj inch in length. When this is to say, the lower electrode must be the negative pole. the case, splashes of liquid are carried up, and the noise of the spark
the
louder. " The best gelatino-bromide emulsion plates are used for photoThe image is developed by pyrogallol, or hydroquinone graphing. and ammonia. Hydroxylamine hydrochlorate, and caustic soda with
is
is also a good developer." sensitiveness of the ordinary gelatino-bromide plate only begins in the green about the /' lines, although it extends into the
little
potassium bromide,
The
NMLH
FED
CBA
Visu.-il
intensity.
Gelatino-bromide (undyed).
Ammoniacal
eosin.
Animoniacal erythrosin.
Ammoniacal
rose-Renixal.
Ammoniacal cyanin.
{From Thorpe's
Of the yellow and red lines desired to photograph in this region, special plates must be employed, made by staining the plate with some dye. Fig. 109 shows the effect of various dyes in extending the sensitiveness of the plate into the yellow and red. By the use of coloured transparent screens, to reduce the action of the blue part of the spectrum, it is possible to photograph as far as the line C on ordinary unstained plates with a long exposure.
far
spectrum
beyond the
visible violet.
it
no trace
is
obtained.
If
is
No single staining material is known which will give a correct representation of all the colours of the spectrum, but by the simultaneous use of stained plates and properly chosen screens, a very fair approximation to the desired result can be attained. Vogel first, in
1873, employed aniline dyes to render the plates sensitive to the green,
The
yellow,
Cottcave Grating.
Photography
of the Spectrum.
127
and red rays. Plates stained with ccerulein,'"' and used with a screen of chrysoidin (a thickness of 1 1 mm. of an alcoholic solution of the strength i 12000), enables the whole spectrum, from the ultra:
violet
up
from
Erythrosine f plates are sensitive to the whole of the spectrum D to H. These are the ordinary " orthochromatic " plates. They are especially sensitive to the green and yellow light. Cyanin J increases the sensitiveness of the plate to orange and red in a marked degree ; whilst the plate possesses less sensitiveness to green. These plates are also useful with longer exposure for the ultra-violet rays. The cyanin bath is made by taking 2 c.c. of an alcoholic solution (x 400) of cyanin with 100 c.c. water and ^ c.c. ammonia. The plates should be prepared and developed in absolute darkness. Congo-red and benzo-purpurin increase the sensitiveness for the region from the green line E to the orange beyond D. The bath is made by adding to 100 c.c. water i c.c. ammonia and 2 c.c. aqueous solution of the dye (i 400). Of plates which may be purchased ready dyed may be mentioned the Cadett " Spectrum plates." These must be developed as far as
: :
possible in absolute darkness ; metol is recommended as the best developer. There are also the plates prepared by the Lumiere Company, which are of three kinds the orthochromatic plates,
:
and yellow ; the B orthochromatic, sensitive to yellow and red and the C orthochromatic, or panchromatic, plates, sensitive to green, yellow, and red. These last must be developed in darkness ; the A plates may be developed by feeble red light, and the B plates may be developed by feeble green hght. It may be noted that the dye substances used to extend the
sensitive to green
;
sensitiveness of the plate towards the red also not unfrequently improve the qualities of the plate for the blue and ultra-violet portion of the spectrum, inasmuch as the sharpness of the lines is increased.
intended for photographing the blue portion, the quantity of the (coerulein, for example) may be increased. The fullest information on the subject of dyeing plates for spectrum photography will be found in the valuable researches of Eder and Valenta, entided " Beitriige zur Photochemie und SpectralIf
dye
analyse." In Fig.
no are shown the spectra of cadmium, zinc, lead, thallium, and tin, from about 2770 to 1940, as photographed by Professors Eder and Valenta, copied by permission from their " Beitrage zur
* Ccerulein is the colouring substance obtained by heating gallein with sulphuric acid ; it does not dissolve readily, and is best used in the form of the compound ccerulein-sodium-bisulphite, which is soluble in water. The plates are stained by bathing them for three or four minutes in a solution of 0-02 grms. of the ccerulein-bisulphite.in 100 c.c. of water with 8 drops of ammonia, and allowing the plates to dry. the potassium compound of tetraiodo-fluort Erythrosine, or iodo-eosine, is The bath must be highly dilute (one in ten thousand) and ammoniacal. esceine.
X
Cyanin
is
chinoline blue.
128
All Introduction
to
the
These photographs are of the In spark-spectrum obtained with a large Rowland concave grating. wave-length the has left the line on strong the spectrum cadmium the 2748, the next strong line is 2573, then three lines al 2329, 2321,
Pholochemie und Spectralanalyse."
'
4i.
p:
lines at 2288 and 2265, and then two at 2194 and The original photographs show many more lines, hardly to 2144. be discerned in the reproductions. In the zinc spectrum the lines to which numbers are attached are
Photography
of the Spectrum.
129
I30
Afi Introdiiction
to
the
which the
of 2771, 2756, 2658, 2502, 2138, 2064, 2061, 2025, 1967, and 1947, first four are strong Hnes. In the lead spectrum we have marked 2663, 2614, 2577, 2476, 2393, 2203, and in the thaUium spectrum 2768, 2709, 2580, 2530, 2469, 2452, 2379, double 2298 and 1964. Fig. 1 1 1 shows the spark- and arc-spectra of cobalt and copper. In the spark spectrum of copper the lines marked are 2769, 2766, 2713, 2689, 2618, 2369, 2303, 2294, 2276, 2247, 2242, 2228, 2210,
2199, 2192, 2189, 2149, 2123, 2104, 2055, 2025, and 1999. In Fig. 1 1 2 we have the spark- and arc-spectra of iron and nickel these photographs show the accuracy with which an enormous number of closely packed fine lines can now be recorded by the photographic method. In the iron spectrum the lines whose wavelengths are marked are 2767, 2747 and 2746, 2720, 2689, 2664, 2628,
;
2611, 2599 and 2598, 2585, 2563, 2534 and 2533, 2493, 2454, 2439, 2413, 2399. 2358, 2373, 2332, 2327, and 2280. In the nickel spectrum the lines marked are 2593, 2576, 2511, 2484, 2473, 2437, 2416, 2375, 22877 and 2287-2, 2270, and 2264.
Absorption Spectra.
It has already been remarked that some substances are best All so-called recognized by observation of their absorption spectra. transparent substances absorb some light, even air and colourless glass, and, as we have seen, the photography of the spectrum is But in the case of these colourless interfered with by this absorption. substances, the absorption may be said to be general, as opposed to the selective absorption of such substances as those of which the These are obtained by placing spectra are shown in Figs. 32 and 33. the solution to be examined in a wedge-shaped glass trough in front of the slit of the spectroscope, and using daylight or some other source of white light. The thickness of the solution increasing in a regular manner from the edge of the glass cell, we obtain a series of spectra corresponding to different thicknesses of the absorbing material, and we observe that certain colours of the spectrum are acted on more quickly than others. Thus, a solution of chromium sesquichloride, if dilute, is of a green colour, but a strong solution, or a greater thickness of solution, transmits only purple or red light. In the same way glass coloured with cobalt oxide is blue in a small thickness, but if many pieces of glass are placed together only red light can make
its
way through.
A very dilute solution of potassium permanganate shows five dark absorption bands in the green, as shown in Figs. 27 and 32. But a very slight increase in the thickness of the layer of liquid or in the strength of the solution obliterates these bands, and only a broad band of blue light and a narrow band of red light are transmitted a further increase in thickness extinguishes the blue band, and only deep red light passes.
Photography
of the Spectrum.
131
Bfea^nfe
IB!!-
132
Afi Introduction
to
the
Fig. 113 shows the absorption spectra of chlorophyll (No. 2) (the green colouring matter of leaves), of blood (No. 3), and of the coloured vapours of iodine (No. 4), and of nitrogen peroxide (No. 5) (nitrous acid). These two gases, as well as other coloured gases, such as bromine vapour, chlorine, and others, are remarkable for the regularity and intricacy of the banded spectra which they give by absorption at low temperatures. The absorption spectrum of iodine vapour is altogether different from the emission spectrum shown in Fig. 27. These differences are no doubt due to different complexity of the molecules in the Iodine dissolves in different states of the same absorbing substance.
Fig. 113.
carbon disulphide or chloroform to form a violet solution of nearly the same colour as the vapour, but these solutions do not give the same banded absorption spectrum as that of the vapour. The molecule of iodine in the solution is, no doubt, more complex than the molecule in the cool vapour, and this, again, is not so simple as the molecule in the electric discharge. Blood possesses so distinct an absorption spectrum, that it is possible to determine with certainty that a reddish fluid, or the stain produced by it, is really blood, even when the microscope fails to give certain evidence. Blood owes its red colour to a substance called hasmoglobin, which is a compound of an albuminoid substance with a red colouring substance containing iron, known as haematin. In arterial blood haemoglobin is loosely combined with oxygen, which is easily removed by treating the blood with reducing agents, such as ammonium sulphide. In cases of
Photography
of the Spectrum.
133
poisoning by carbon monoxide (e.g. from water-gas), the haemoglobin enters into combination with carbon monoxide instead of oxygen, and this compound as will be seen possesses its own characteristic absorption spectrum. Even a very dilute solution of blood gives two characteristic absorption bands between D and E, whose wave-lengths are 5774 and 5390. On addition of ammonium sulphide, two strong bands of reduced haematin at 5570 and 5240 are produced; the second of
BCaD
E&
134
-''
Fig. ii6.
Fig. 117.
135
Figs. 116 and 117 show such a micro-spectroscopic eye-piece. The prism is contained in the tube A. Below the prism is an achromatic combination, having an adjustable slit between the two lenses, the upper lens being furnished with a screw-motion to focus the slit. are milled heads C and which adjust the slit. is an arrangement for holding a small tube, so that the spectrum of its contents may be compared with the spectrum of the object on the stage of the microscope. I is a mirror for throwing light through this, and E a screw regulating the amount
Fig. 118.
(From
I.
[[''lifts s
IT.
III.
,, ,, ,, ,,
IV.
V.
VI.
a phylloxanthin derivative.
,,
The e.xact of light which is admitted to form this second spectrum. focussing on the lines of the spectrum is done by means of the milled head B. A Chlorophyll is insoluble in water but soluble in alcohol. solution made by digesting green leaves with alcohol, if strong, absorbs nearly all the spectrum only red light being transmitted; but with a more dilute solution absorption bands are developed in Addition of acid changes the colour to dull yellowish the green. green or olive, and from this acid solution Fremy has separated two other colouring substances, which are called phyllocyanin and
136
An
Introduction
to
the
phylloxanthin. Phyllocyanin form a blue solution in hydrochloric Phylloxanacid which gives a spectrum with five absorption bands. thin is yellow, and has a spectnmi with four absorption bands. These spectra are shown in Fig. 118. The spectroscope is often of value in identifying particular dyes or colouring substances and in tracing them even when mixed with
BC
E 6
F
Alizarine, in alcohol.
Alizarine, with
ammonia.
Aqueous
solution, with
ammonia.
Anthraquinone sulphonic
ous caustic potash.
Purpurlne
dotted).
Purpurine
in alcohol,
with ammonia.
Purpurine
in alcohol,
Purpurine,
dotted).
with
alum
(strong
solulion
ammonia.
Fig. 119.
[From Thorpe
s "
Other coloured substances, and thus in detecting adulteration. The colouring substances in madder, or obtained from madder, have
which the most important are alizarine and purpurine, both now manufactured artificially from the hydrocarbon anthracene. From anthracene, anthraquinonedisulphonic acid is made, and this by treatment with caustic potash gives dioxyanthraquinone, or aUzarine. The absorption spectrum of alizarine is marked
characteristic spectra, of
Photography
of the Spectrum.
;
37
by two bands, one about D and the other near C and purpurine is marked by different bands in the neighbourhood of F and E. The genuine red colouring substance of wine, derived from the skin of the grape, has a spectrum quite different from that of magenta or other dyes sometimes used in sophisticating " wine." The
BC
E5
F
Cochineal, in water. Cochineal, in alcohol, weak (strong solution dotted).
Corallin, in alcohol.
Corallin, in water.
weak
solution
(strong
solution
dotted).
Eosin, in
amyl
alcohol.
Safranine,
dotted).
weak
(strong
Turmeric, alkaline.
Dragon's blood.
Fig. 120.
[Fyom Thorpes
''
detection of such adulterations does not depend only on the differences between the spectra themselves^ but also upon the series of alterations in the spectrum brought about by treatment with various reagents. Cubes of glycerine jelly soaked in the wine to be tested are not much stained if the wine is genuine, but many of the dyes penetrate the jelly throughout, and a slice cut from the cube may be
may be
CHAPTER
XI.
RELATIONSHIPS BETWEEN THE DIFFERENT LINES OF A SPECTRUM, AND BETIVEEN LINES OF THE SPECTRA OF ALLIED ELEMENTS.
of the most interesting subjects of speculation is the inquiry between the different lines of the same spectrum, and of relationships between lines of the spectra of allied elements. One cannot but be impressed by the beautiful rhythmical arrangement of the lines forming such groups as those of
into the existence of relationships
Oke
the carbon spectrum (Fig. i6), of cyanogen (see Fig. 51, lower strip, and B in the solar spectrum between 7 and 8), and of the groups (see Figs. 56 and 57), which seem to be repetitions of each other. The lines we observe are due to vibrations in the ether of definite period, and the period is in some way determined by the particular atoms or molecules of the substance, which as we say " give that
spectrum."
Even in the early days of spectrum analysis it was felt that there must be some connection between the vibrations to which the different lines of a glowing gas are due, and the earlier attempts at A tracing this connection were based upon analogy with music.
pianoforte string, for example, is capable of giving not only its fundamental or lowest note but also a series of overtones ; the octave, of just twice the frequency the twelfth, or fifth of the octave, of just three times the frequency, and so on. In Fig. 25 we have shown the spectrum of hydrogen in a vacuum tube of four lines, known as Ha in the red, H/8 in the green, and Hy and H8 in the blue; and in the spectra of the stars (see Fig. 73) we have a con;
tinuation of the spectrum, so that the complete spectrum is made up of a series of lines getting closer and closer together as we advance into the ultra-violet. Professor Johnstone Stoney endeavoured to
explain this rhythmical arrangement of lines on the theory that they were due to overtones of a low fundamental vibration, and he showed that Ha, H/?, and H8 might be the twentieth, twenty-seventh, and thirty-second harmonics of a fundamental vibration of wave1 length 131 27 7 4. But there was no place found for Hy in this arrangement, nor does it explain the series of stellar lines of
Relationships in Spectra.
139
hydrogen. It is now generally admitted that such theories will not account for the facts, and a definite end was put to such speculations by Professor Schuster, who showed that the coincidences observed were not more numerous than ought to be the case as the result of chance, as we are taught by the theory of probabilities. The first striking success in the attempt to explain these regularities in spectra was obtained with the hydrogen spectrum by Professor Balmer, in the year 1885, who showed that the wave-lengths of the hydrogen lines could be calculated by a very simple formula, viz. the oscillation frequency
= 27418-75(1
where
--J
5, 6, 7, 8, 9, etc., in
is
to
have given
to
it
the values 3, 4,
hydrogen is given in the accompanying table, seen the remarkable agreement between observed and calculated values. It will be noticed that the lines for which values are calculated from the formula are the brightest lines, but there are many other lines of less intensity which do not fall into this series. These belong to what is called the secondary or " compound-hne-spectrum " of hydrogen. It should be noted that Professor Trowbridge is of opinion that the " four-line " spectrum of hydrogen is really due to water-vapour and not to the element hydrogen, and that the true spectrum of hydrogen is to be found in this compound
succession.
of
spectrum.
140
An
Introduction
to
the
Relationships in Spectra.
141
142
An
Introduction
to
the
It is generally easier to trace such regularities in the ultra-violet portion of a spectrum, because with decreasing wavelength the lines of a series approach continually closer together, and thus their character of symmetry strikes the eye more easily. Again, we can hardly expect all the lines of a spectrum, or indeed any large proportion, to fall under any one formula, for in a luminous gas we
any connection.
have
particular spectrum, probably, does not correspond to a definite temperature, but is a mixture of the spectra of different temperatures, in which the spectrum of the mean temperature predominates. AVe
cannot assume that all the atoms of a molecule vibrate in precisely same manner, and if they do not, then each atom may give rise to a separate series, which will require a separate formula for its Again, of lines belonging to temperatures far rerepresentation. moved from the mean temperature, only the most intense may be visible, and it will not be possible to find any formula connecting them. Thus in the spectrum of hydrogen it is only the strongest which can be represented by Balmer's formula. In the case of some elements, such as the metals of the alkalies, the spectrum seems to change little with change of temperature, and
the these are therefore the easiest to deal with. In the arc-spectrum of lithium twenty lines have been observed, and Kayser and Runge find it possible to account for all of the hnes by means of formula;. The lines fall into three series, which are called respectively the principal series, the first subordinate series, and the second subordinate series.
Principal series First subordinate series
.
Second subordinate
series
The formulae for the three series are:^ = 43r8473(i - 3-o6688w-= - 2c,-240i2//i-') = 2858674(1 - 3-S3479-= - o-o646o96?-') = 28666-69(1 - 4-26945OT-= - 8-o8255OT-^)
The
calculation
of lithium
following table shows the agreement between the results of and the observed wave-lengths of the lines in the spectrum
:
Value o(
111
and
seri
Calculated wave-length.
6600 '08
6103-77 4972-11
460-2-37 4-273-44 413-2-44
Relationships in Spectra.
In
this table the principal series is
143
first
marked
P, the
subordinate
series
series B.
The wave-lengths
The
agreement between calculated and observed values is very satisfactory except for the red line. It is observed that the lines of any one series agree
in character
thus the lines of the ; principal series are sharp and easily reversed, those of the first subordinate series are diffuse and reversible, and those of the second subordinate series
are diffuse on the red side,
and are
ultra-violet
fe
The
the secondary series seems to indicate that they are due to the vibrations of more complicated molecules, belonging to a lower temperature.
We
expect a whole series to be absent as the result of dissociation under particular circumstances, but it is not likely that we shall have some lines of a series while the brighter
lines of that series are absent.
may
In the spectra of the other metals of the alkalies, viz. sodium, potassium, rubidium, and CEesium, similar relationships are found ; the lines in each case being divisible into a principal series and two subordinate series, in one of which the lines are sharp, and There is, howin the other diffuse. ever, this difference between these spectra and that of lithium, that instead of single lines, as in lithium, we have pairs of lines in sodium, potassium, rubidium, and csesium. The arrangement into series in the case of sodium is shown in Fig. 121,
144
-"^
Introduction
to
the
where P denotes the principal series, D the diffuse, and S the sharp subordinate series. The formute are as follows
:
ioA-' Principal series Subordinate sharp fio'x-' \ioA-' series Subordinate diffuse f lo'^-' \io'\-' series
. .
= = = = =
- 3-12939^-^ - I9'33950'~') - 4-9I774ot-^ - 8-o6i032 ') - 4-9I485ot-2 - 8-05882OT-') - 4-49905^-= - o-I46997ot-') 4-49905"'- o-ll9447;~')
D lines correspond to wi = 3 in the principal series. appears that the double lines of the principal series get closer and closer together as the value of m increases, but that in the two subordinate series the double lines are always at about the same distance The same rule applies to the double lines in the spectra of apart. These formulse account for all the lines in the other alkali metals. the arc-spectrum of sodium with the exception of a pair at 5675-92, 5670-40. Similar relationships are found to hold in the case of potassium there is a principal series of pairs of lines extending far into the ultraviolet, and two subordinate series, also of pairs of lines, in the visible portion of the spectrum. The red and blue lines of the flame-spectrum belong to the principal series, and correspond to ;;/ = 3 and m 4. The formula given by Kayser and Runge for potassium are as
The Fraunhofer
It
follows
(10"^-'
Prineioal imcipai series 1
First
. .
-^jq,^-,
= ^ = = = =
3508655(1
35085-90(1 22021-83(1 22077-11(1 21991-24(1 22050-32(1
20434ot~-
I7S22l6-')
17-71746OT-') 2-83837w;~')
2SgSo'jm~')
5
5-19167^"-
o54Ioot~')
5-04922w~')
inspection of these forniulse would suggest that probably, in the series, the coefficients of ;~^ and m~' should have for the two lines of a pair, which would give the succession of pairs of lines at the same distance apart ; while in the principal series, if the first and second coefficients had the same value, then the width of the lines apart would depend on the difference between the values of the coefficients of m~*. The distance apart of the lines of the pairs would then decrease with increasing values of m, and would be inversely proportional to the fourth power of m, which is what is found to be very nearly the case. All the fortyone lines of the arc-spectrum of potassium are accounted for by the above formute, with only small differences between the calculated values and those observed, but the formulae do not give accurately the wave-lengths of the infra-red lines observed by Becquerel and by
An
Snow.
Rydberg
prefers another
formula, viz.
lo^'A.^'
where a stands for the convergence frequency, the value 109721-6 and = 0-254. Over a certain range of spectrum the two formula, Rydberg's and that of Kayser and Runge, agree with the observations about equally well, but neither will fit the whole range
jj.
= and N has
<?
-.
N -
y,,
Relationships in Spectra.
145
tlie ultra-violet.
is
N
X
in his formula
the
same
Rydberg thinks that the value for all elements (or that the product
and
B, in Kayser and Runge's formula, has always the same value), that the two subordinate series in each element end exactly at
it
the same point. This does not appear to be exactly true, though is very nearly so.
In rubidium
closer together as
The formula
= = = =
33762'ii(i 33765'38(i
37i78i/-= 37i4782~'
i6-65343;-<)
i6-i3i29-')
series
2o939-39{i 21179-38(1
6-42884W-')
6-22299OT-')
Again
it
seems that
all
quite similar relationships a principal extending into the ultra-violet and two subordinate series.
:
The
formulae are
fiox-'
tio;.-'
'
= 3i50i-56(i-3'97050OT-2- l5-5Sio7/-') = 3i46578(i-3-9i498;-^-i47646i'-') = i9743'25(i 6-22335ot-= i5-4902ot-<) lo'A. First subordinate series io'A~' = 20295 22(1 -6'055i7/~- i5-6oioot~') Second subordinate series The spectra of magnesium, zinc, and cadmium are characterized
PrinciDal series lrmcipalser.es
.
.
by the presence of triplets, repeated over and over again close triplets in magnesium, wider in zinc, and still wider in cadmium. The Fraunhofer b group is one of these magnesium triplets, shown in Fig. 51, and in Fig. 23 the same is seen at 75, and in the zinc spectrum a triplet at 95, 100, and 102 ; and in the cadmium spectrum a triplet at 80, 95, and 103. Kayser and Runge arrange the triplets of magnesium in two series, which they think to be both subordinate series, and find that the oscillation frequencies can be calculated very nearly by means of
;
lio*A-'
First series
.
.
lo'A.-'
io'A~'
jio'A."'
Second
series
io*A-' (io'a~'
I30398;-= I30398?~=
I30398~^
125471OT--
125471^-=
i2547i?~^
By means of these formulae, 39 lines out of the 56 lines in the arcspectrum of magnesium are accounted for, with very small differences between the calculated and observed values, except in the red, where It will be seen that, accordthe differences amount to i per cent.
ing to these formulae, since the coefficients of w/^^ and nr^ are the same for the three lines of the triplets in each series, the spacing of the lines in the successive triplets will be always the same. L
146
Ah
Introduction
to
the
FirsL series
<io'A.-'
I
io*A~'
iio'A.-' 10"^-'
io*A."'
= = = = = =
961696///"'
96l696/-'
961696///-' 346097///-' 346097///-'
346097^/-'
For strontium, Kayser and Runge Lehmann has discovered a second series.
iio'A-i io\-'
lo'A."'
iio''A.-'
find
io'a"' io*A~'
_ = = = = =
837473^-'
837473///-' 837473///-' 301000///-'
301000///-'
- 3oiooo/-'
In the spectrum of barium, with 162 lines, Kayser and Runge did not succeed in finding any formula. In the spectrum of zinc, Kayser and Runge find two series of triplets which can be represented by the formula
= = = = =
1236125///-'
1236l25/-<
1236125///-' 632850///-' 632850///-' 632850?//"'
The
for 36 lines out of the 73 lines in the arczinc spectrum shows an extraordinarily strong line
that it is "the most This line, and several other strong lines, are not accounted for by the formute, so that the whole spectrum is far from being reduced under any one law, as was
2i38'3; of which
Professor
Crew remarks
the case with the simpler spectra of For cadmium the formute are
First series
tl"ie
<
Second
series
loi^A"'
555137/K-'
555137?//-'
These formula; account for 36 lines out of the 74 lines in the arcspectrum of cadmium. We have again several strong lines unaccounted for, e.g. one at 2288"io. The spectrum of mercury appears to be more complicated than those of zinc and cadmium. Kayser and Runge account for 25 lines out of some 92 lines in the arc-spectrum by the following formulae !io*A-' = 40159-60 - 127484///"= 1252695///-' lo'A"' = 44792-S7 127484///-= 1252695///-'
:
Second
series
127484///-= 126361///"=
1-26361///"=
--
1252694TO-'
613268///"' 613268///"'
126361///"=
61326S///-'
Relationships in Spectra.
147
are
at
In the mercury spectrum, as in those of cadmium and zinc, there some strong lines not accounted for, in particular the strong line
253672. The arrangement of lines in series in the spectra of magnesium, calcium, strontium, zinc, cadmium, and mercury is shown
in Fig. 122.
'S^
Cl
In the spectra of copper and silver, Kayser and Runge find that the lines occur in pairs with a constant difference in oscillation frequency, as shown in Fig. 123. They find that there are in each case two series of pairs, for which the formulae are, for copper
148
An
Introduction
to
the
First series
I lo'A.-'
\io'A~'
Second
series
= 3i59f6 - I3II50OT~ = 3i840'i i3n5om~ fio'A""' = 3i59i'6 i248o9ff;~ \io'\~' = 3i840'i i248o9;~
= 30712-4 3i633'2 = 3o696'2 = 3i6i7'o
for very 13062IOTi3o62iffj-
440582/-' 440582n~'
and
for silver
First series
flo'A
1093823///-' 1093823///-'
Second
fio\-i
series
123788OT
123788//;
394303///-'
(io"a~'
394303'""'
copper and
silver.
Fig. 123.
{From
W'iitls's
''
Dictionary of Chernisfry."
In the spectra of the elements aluminium, indium, and thallium, find in each case two subordinate series of pairs of lines. Their formulas are, for aluminium
First series
Second
series
(lo\(lo'A(lo'AIio'a-
156662///156662///127527///127527///-
indium
First series
Second
series
thallium
First series
Second
series
41542-7
-1
~ ~
1265223///1265223///790683///-'
790683/n-'
For the spectra of tin, lead, arsenic, antimony, and bismuth, no formulae have been discovered, but certain regularities have been noted. In the spectrum of tin the lines 3801-16, 333071, 2850-72, 2813-66, 2785-14, 2779-92, 2594-49, 2571-67, 2524-05, 2495-80,
2408-27, 2358-05 and 2317-32 are repeated in the lines 3175-12, 2840-06, 2483-50, 2455-30, 2433-53, 2429-58, 2286,79, 2269-03, 2231-80, 2209-78, 2141-1, 2100-9, 2068-7, that is to say, the first
thirteen lines are respectively at the
the second
Relationships in Spectra.
thirteen lines
149
the scale of oscillation frequency, of these ; the seventh, ninth, and eleventh occur again further on. In lead also ten lines are repeated with a constant difference, and eight of these are repeated again. The same occurs with eight lines in the spectrum of arsenic, and in antimony a set of lines seems to be repeated no less than five times. Another subject of inquiry of much interest and importance is that of a possible connection between the spectra of allied elements. The study of chemistry leads us to associate certain elements as forming natural families ; e.g. fluorine, chlorine, bromine, and iodine form a well-marked group, so also lithium, sodium, potassium, rubidium, and caesium ; calcium, strontium, and barium ; magnesium, zinc, and cadmium ; phosphorus, arsenic, and antimony, and so on. MendelejefiT's periodic law is an attempt to connect the properties of the elements with their atomic weights that is, of the smallest masses of the elements known to the chemist.
first
when reckoned on
five
Groups
I50
A/i
Relationships in Spectra.
151
1891, confirmed this statement, and pointed out many connected by the formula already given. They further call attention to the fact that the differences in oscillation frequency in the corresponding lines of the triplets increase with increase of atomic weight, and that they are nearly as the squares of the atomic weights. They also remark that in the spectra of potassium, rubidium, and cjesium, the differences in oscillation frequency between the close pairs of the subordinate series are nearly proportional to the squares of the atomic weights. These relationships are most clearly exhibited by plotting squares of atomic weights as ordinates and oscillation frequencies as abscissae, that is, by making a diagram on squared paper (like Fig. 46), marking off a scale of oscillation frequencies on the bottom edge of the paper and a scale of squares of atomic weights along the left-hand edge, and marking down the positions of the lines of the spectra. Ramage, who has discussed these relationships in a paper communicated to the Royal Society {Proc. Roy. Soc, 1901), has given such a diagram. He points out that the lines connecting corresponding points are nearly straight in the case of potassium, rubidium, and caesium ; and that the straight lines drawn through corresponding members of doublets and triplets in such cases as those of zinc and cadmium intersect on the line of zero atomic weight. In cases where this law holds good we can, then, determine the atomic weight of an element from that of another, or from those of two others, if only we can be sure that the lines chosen are really Thus, for homologous, or do really correspond to each other. example, we may calculate the atomic weight of zinc from that of cadmium (ii2'4) by means of the corresponding triplets in the spectra of these two metals. The differences between the first and third lines of the triplets are as follows
in
Runge,
more
triplets
Cadmium.
152
An
Introduction
to
the
though not so exactly between similar relationship holds the pairs of lines in the subordinate series in the spectra of sodium, The mean differences between potassium, rubidium, and csesium. the pairs are given by Kayser and Runge as follows
:
Relationships in Spectra.
53
dependently by Professor Rydberg and by Professor Schuster. If we call the slowest vibration of a series the " fundamental " vibration, i.e. the one for which the value of n in Kayser and Runge's formulae is 3, the convergence frequency of the principal series can be calculated approximately, at least from the fundamental vibration and from the convergence frequency of the subordinate series. The convergence frequencies of the two subordinate series are nearly the same; in the case of lithium they are 28586'7 and 286667. If to the mean of these numbers, viz. 28626'7, we add the fundamental oscillation frequency 14907 "i, we obtain 43533"8, which is nearly the value of the observed convergence frequency of the principal series, viz. 43584"7. The rule may conveniently be expressed thus: CFp = CFg + OFp, where CFp means the convergence frequency of the principal series, CF, the mean convergence frequency of the subordinate series, and OF, the oscillation frequency of the fundamental vibration.
The
far the
law
.
how
Sodium
35086-2 33763'5
I3'i
We cannot make the calculation in the case of csesium, because the fundamental vibration lies in the infra-red, where no lines had been observed in 1896 when this rule was enunciated but Professor Schuster ventured upon the prediction that infra-red hnes of caesium would be found with wave-lengths 8908 and 8518: possibly these It are the lines since observed by Lehmann at 8949'9 and 85277. will be noticed that the differences between the calculated and observed values decrease with increasing atomic weight. No principal series is known in the other elements, but calculating from this rule we may form some idea as to whereabouts the
;
principal series
would terminate, and we find that in many cases it would lie far in the ultra-violet. Using the first lines of the triplets in magnesium, calcium, zinc, etc., we obtain the following approximate values of the convergence frequency of the principal series
:
Magnesium
Calcium
Zinc
59 '27
50268 63731
Cadmium
Mercury
60459 5^530
consideration of these numbers seems to show that magnesium not go with zinc, cadmium, and mercury, but that it fits in best with calcium, strontium, and barium ; and that lithium and sodium go together, but not with potassium, rubidium, and caesium.
will
154
Am
hitrodnctioH
to
the
Study of Spectrum
A na/ysis.
We
in the case of the metals of the alkalies, and and cadmium, the distances apart of correspond-
But this ing lines are nearly as the squares of the atomic weights. It may be, however, that in other groups of related elements the distances apart of the lines of a pair (or triplet) are proportional to the atomic weight raised to some other power than two exactly. Thus for copper and silver the
rule certainly does not hold universally.
power would be about 2*4765, for magnesium, calcium, strontium, and barium, i'62i4, and for aluminium, gallium, indium, and
thallium, 2'ii27.
Runge and Precht express this by plotting as ordinates the logarithms of the atomic weights, and the logarithms of distances apart as abscissa ; then the points representing chemically related elements lie on the same straight line. This appears to be nearly (but again only nearly) true in many cases, and from this relationship we can obtain a formula for calculating the atomic weight from the separation of the lines of a pair (which we will call x).
aluminium group we have the formula the the number whose logarithm is the number we obtain by multiplying the logarithm of the separation by 0-47327 and adding 0-47 185 ; or we may write the formula atomic weight
Thus
for
the
is
atomic weight
log~^ (0-47185 0-473327 log .v). calculated values for the atomic weights
We
:
Boron Aluminium
Gallium Indium Thallium
....
1083
27-65 71-23
instead of
11
27-1
70-0
14-0 2041
'3'54
206-04
is
log-' (0-54057
8'8
+ 0-49285
instead of
log x)
This gives
Beryllium Zinc
9-1
Cadmium
Mercury
,, ,,
log-' (0-45705
+ 0-57223
log x)
Magnesium
Calcium
Strontium
....
silver
Atomic weight
Lithium Copper
Silver
log-' (0-83405
+ 0-40378
log x)
Gold Mercury
6-S2 instead of 7-03 63-6 63-6 ,, 107-93 107-95 '> 196-7 197-2 ,, 2o6-2 ,, 200-O
Relationships in Spectra.
155
Mercury seems to fit in here better than with zinc and cadmium lithium is included by putting log : o on the supposition that the lines of lithium are doubles with an infinitesimal separation. For the metals of the alkalies the formula
Atomic weight
gives
log-' (o'65284
+ 0'53869
,,
,,
log x]
Sodium
Potassium
20'82 instead of
23 '05
39' 15
Rubidium
Cf^sium
.....
3962
5 '02 i33'94
4'5.
85 '4
..
133
would seem to confirm our previous conclusion that lithium and sodium will not go with potassium, rubidium, and caesium ; and that magnesium must be placed with calcium and strontium and not with zinc and cadmium.
results
These
ttie
A
same
lines, as
comparison of the accurately measured wave-lengths of the observed in the solar spectrum, and as produced in
the spark and arc, reveals in nearly every case a slight difference, the lines in the arc and spark having a slightly greater wave-length than their corresponding reversals in the solar spectrum. The difference is in most cases extremely small, but there can be no doubt of its reality, and the matter becomes one of importance in view of the use of the Fraunhofer lines as standards of wave-length. If we use a line in the spark- or arc-spectrum for the purpose of comparison, and as a standard to measure other lines by, it is most important to know if its wave-length differs, even in the smallest degree, from that of the same line seen as a Fraunhofer line ; and it may become necessary to look to the vacuum tube for lines of The unalterable wave-length to serve as standards for the future. following table showing the wave-lengths of some calcium lines in the arc, in the spark, and in the sun, will show the amount of difference
which may
exist
In the sun.
156
An
Introduction
to
the
be altered by changes
An
in the amount of material present in the arc. increase in the density of the vapour producing the lines in general increases their width some broadening symmetrically, others more on the one side or the other, but generally most towards the Again, some lines easily become reversed ; others are reversed red.
:
with difficulty or not at all. The effect of pressure upon the exact position of lines has been put to the test of direct experiment by Humphreys, by producing the arc in an enclosed space, into which air was pumped so as to obtain The result of these high pressure up to 13 or 14 atmospheres. experiments was to show that increase of pressure in almost every case caused a shift of the line towards the red end of the spectrum, the amount of which was proportional to the increase of pressure and to the wave-length, but that the amount of shift was very different for different elements, and even different for different sets of lines in the spectrum of one and the same element ; thus in the alkalies the lines of the principal series and of the subordinate series are unequally shifted these shifts (for equal wave-lengths) are to each other approximately as I 2 4 for the principal, first subordinate, and second
;
'
subordinate series respectively. and are found to be affected by pressure The calcium lines only about half as much as the g calcium line at 4226-9 ; the metals
The lines of banded the largest shift. cyanogen and of alumina, do not seem to
be affected by pressure. The displacement of lines in the sparkspectrum towards the red is beUeved to be due to a high pressure in the metallic vapours thrown off from the electrodes with a high a velocity estimated by Professor Schuster velocity in each discharge To account for the to be as high as 2000 metres per second. displacement of the spark lines on the basis of the measurements of pressure in the arc just described, a pressure of from 24 to 30 atmospheres must exist in the metallic vapours in the path of the That a stream of sparks in a partially exhausted space does spark. cause a sudden increase of pressure was shown directly by De la Rue and Miller so long ago as 1879. The possibility of any difference between the wave-length of a line as produced in the arc, and that of the corresponding Fraunhofer line, was not taken into account by the late Professor Rowland in constructing his " Table of Standard Wave-lengths," which is given in Any such differences were supposed by him to be the Appendix. due to accidental movement of the apparatus when changing from the spectrum of the sun to that of the arc, and the wave-lengths of metaUic lines read off on the " arc " photographs were corrected by
iron lines on the plate, so as to bring them into agreement with the measurements made in the solar spectrum. This is probably, in part, the explanation of certain differences between the numbers obtained for some bright cadmium and mercury lines by Messrs. Michelson and Fabry and Perot by the use of methods of interference not employing the concave grating, and Rowland's
Effects of Pressure
and
Density.
157
lines. By counting the number of interference fringes in the length of the standard metre at the Bureau des Poids et Mesures in Paris, formed in the red cadmium line and also in the green and blue lines, Michelson has obtained results which are probably correct to a few units in the thousandth place, and his result is thus more accurate than the best determination of wave-length made with any grating. Michelson's values for these three lines are whereas Rowland's values are 6438'472, 5o85'824, and 4799"9ii 6438-68o, 5o86'ooi, and 48oo'o97."' It would appear that the values given in Rowland's Table for the standard lines in the arc have been "corrected" in a manner which cannot now be followed. Messrs. Fabry and Perot have measured with their apparatus the wave-lengths of thirty-three solar lines. Their results are given in the following table, and Rowland's values for the same lines
;
:
158
values
An
Introduction
to
the
accuracy than
and when we require greater becomes necessary to take into account variations due to pressure, density, and other conditions of the source of light. It seems to be even more important to have a generally accepted
may be
this, it
standard than to have great accuracy in the absolute values. We have seen that the spectrum of any particular metal is liable to considerable variation, according to the mode of its production, the temperature, and pressure of the luminous gas, the quantity of Fig. 20 material present, or the density of the luminous gas, etc.
must be
clearly
For example, the longest lines are not necessarily the strongest lines. the magnesium line 4481 is strong but short in the spark, and is absent from the arc (see Fig. 135); the blue lithium line at 46o2'4 In the spark-spectrum of lead, the line 4058 is is strong but short. one of the longest lines, but quite faint \ it is, however, the strongest line in the oxyhydrogen flame-spectrum, and one of the strongest lines in the arc-spectrum, and the reader can easily find other If instead of taking the spark between electrodes of the examples.
metal, we use electrodes of platinum moistened with solutions of salts of the metals, we have generally a smaller number of lines, and it is observed that it is the short lines which are missing those which The same thing persist are not the brightest lines, but the longest. is observed in using electrodes made of an alloy ; the spectrum of a metal present in very small quantity, perhaps only as an impurity, is
:
reduced to one or two of the longest lines. Thus cadmium, as an impurity in other metals, will be detected by the presence of the line 50S6, the longest line in the spark-spectrum of cadmium. When lines are reversed, it is always the longest lines which are reversed first. The two aluminium lines between and K, shown in Fig. 20, are the longest, but not the strongest, lines in the sparkspectrum. They are the only aluminium lines reversed in the sun. These observations show that the character of particular lines of a spectrum may be modified by the mere presence of other substances, without chemical action. The dilution of a gas with a large proportion of some other gas has the same effect as reducing the pressure, making the lines due to the first gas sharper, and, on the other hand, an increase in pressure or in quantity of material has the effect of expanding the lines. This may be easily observed with the spectrum of sodium in a Bunsen flame. In a mixture of gases, the molecules of any one gas come into collision with like molecules less frequently, and the effect is the same as if the gas were observed under conditions of reduced temperature or quantity of material. For further iflustration of these points the reader is referred to the interesting papers by Sir W. Huggins reprinted in the Appendix. When we compare together the spark and arc we see that in the middle of the spark we have only the longest lines, and at each pole, where the temperature is the highest and the density of the metaUic vapours the greatest, we have the short lines, whereas in the middle
Nature of
the Spark-discharge.
59
of the arc the density is greatest and the temperature highest. Consequently we get the long lines from the poles and the short lines from the centre of the arc.
The method of long and short lines cannot be used with the concave grating in consequence of its astigmatism that is, that each point of the luminous object has for its image a line. The method
is
only available with prisms and plane gratings. It was observed by Kirchhofif and by Thalen that certain changes could be produced in spark-spectra by inserting a wet string in the circuit containing the Leyden jar and spark-gap, so as to increase its resistance, or by causing the spark discharge to pass through the coils of an electro-magnet, which increases its self-induction. The spark is a complicated phenomenon. When the difference of potential between the electrodes becomes great enough, there occurs a first brilliant flash, giving the air-spectrum ; then the space between the electrodes becomes filled with metallic vapour, through which a series of oscillatory discharges takes place, giving rise to the metallic spectrum. Schenck, studying the effect of introducing self-induction into the circuit upon the lines of the cadmium spark, finds three sets of lines which behave differently. One set of lines (C), extending only a short distance from the poles, disappeared on the introduction of self-induction ; they are none of them reversed, and they are absent
from the arc-spectrum. The second set of lines (B) includes the rest of the " spark " lines that is, lines absent from or weak in the arcspectrum, though prominent in the spark-spectrum. Some of them
extend right across the spark-gap, but they are only strong close to
the poles.
The
effect of self-induction
upon them
is
to cause
them
to shorten or disappear.
The
so-called arc-lines, including most of the lines which are common to both arc and spark. Some are extremely weak or absent in the The effect of self-induction upon these spark, but strong in the arc. lines is to make them approach more nearly to the condition in the arc. They extend right across the spark-gap with a self-induction which shortens the lines of group B close up to the poles, and are no doubt due to metallic vapour. The lines of groups A and B are due to the oscillatory discharge, the lines of group A, while partly due to the oscillations, are also caused by the metallic vapours filling up the gap, which retain their luminosity after the oscillations have ceased. By photographing with a revolving mirror, Schenck succeeded in showing that the " arc " lines in the spark persisted about twice as long as the " spark " lines, viz. about o'oooo4 of a second.
CHAPTER
BAh^D-SPECTRA.
XII.
SPECTRA
OF COMPOUNDS.
have seen that the spectra of the elements usually consist of and that there are some spectra both emission and absorption spectra which appear to be due to compounds the diagrams of
lines,
We
show that the flame-spectra of the chlorides of barium, strontium, and calcium are complex, being made up partly of bands due to the undecomposed chloride, partly to the oxide of the metal, with only one or two lines which can be traced to the metal itself In Fig. 26 we have a band-spectrum and a line-spectrum of sulphur, and we have many other examples. The spectra of " fluted bands,"
Fig.
1
or " channelled spaces," when seen with small dispersion, are generally made up of bands sharply defined on the one side and gradually fading away upon the other. When seen with greater dispersion, each band is seen to be made up of a large number of very fine lines which come closer together as the sharp edge is approached. Of spectra in which the bands are sharp on the side towards the red (or least refracted side) and fade away towards the blue, we have examples in the spectrum of the Bunsen flame (the " Swan " spectrum). Fig. 16 (9), the nitrogen vacuum tube spectrum, Fig. 27 (i), the beautiful green band seen in the spectrum of burning magnesium. Fig. 23 (3), at 89, the bands of cyanogen, one of which is shown in the lower spectrum of Fig. 51 opposite the number 78 of the scale. Band-spectra which show the opposite character, of sharp edges towards the violet and bands fading away towards the red, are more
we have examples in copper oxide. Fig. 17 (1), manganese oxide, Fig. 17 (4), sulphur, Fig. 26 (2), aluminium oxide, bismuth oxide, chromium chloride, copper chloride, gold chloride, lead oxide, and the terrestrial bands in the solar spectrum due to oxygen and water vapour. Figs. 56 and 57, the absorption spectra of bromine, iodine vapour, iodine monochloride, and nitrogen peroxide.
;
common
It seems reasonable to suppose that band-spectra are due to more complex molecules than line-spectra. We know that sulphur vapour at a low temperature has a higher specific gravity than when it is raised to a much higher temperature, and that consequently at the low temperature the molecule is more complex, and also that at
very
high
temperatures
the
dissociated.
Band-Spectra.
Spectra
of Compounds.
i6i
made to discover some law of the dismake up these band-spectra. That there is some law underlying the arrangement no one who studies such photographs as those of the Fraunhofer groups A and B in Figs. 56 and 57 can possibly doubt. In the same way as the lines of one of
attempts have been
tribution of the lines which
Many
Kayser and Runge's series approach closer and closer together up to the convergence frequency, so the lines of a band seem to get closer and closer together till they combine to form the bright edge of the band. Deslandres has proposed a formula of the form
Bw'^, where 711 has in succession the values o, i, 2, 3, 4, 5, and Kayser and Runge have found such a law verified in their measurements of the lines of the cyanogen spectrum forming the bands beginning at 4606-3, 42i6'i, and 3883-5 (see Fig. 51).
A."'
=A+
etc.,
Thiele has undertaken a very laborious research with the object of finding a satisfactory formula, and adopts A. /[(;/ cf\ where c is the same for any series (" the phase "), and n has the values I. 2, 3, 4, 5, etc. is some function of {n cf, and if n be given successively all real integral values will have a maximum value at A = /(O), where the lines unite to form an edge, or " head," and a minimum value at A. =/( oc ), where there should be a crowding of lines together to form a " tail."
"''
Professor Thiele has tested his formula on the brightest band of the " Swan" spectrum, that shown at 76 in Fig. 16 (9), with " heads" at 5i65'3j 5I29'4i 5ioo(?)) 5067, and 5050, and regards the correspondence between calculated and observed values as satisfactory. Professor King t by long exposure on an arc between poles of pure
carbon has discovered some new cyanogen bands at 3465-7 and 3203-8, which appear to be really the " tails" corresponding to the heads at 3883-55 and 3590-52 respectively. The history of the " Swan " spectram is interesting. It was first observed by Wollaston in 1802, but first described by Swan in 1857, and was considered by him to be due to a hydrocarbon. PKicker and Hittorf give a fine coloured representation of it in the Philosophical Transactions of 1865. Attfield came to the conclusion that the spectrum was that of carbon, since it was common to compounds of carbon with hydrogen, nitrogen, oxygen, and sulphur. Morren undertook experiments to prove that Attfield was wrong, but became convinced that he was right. Morren particularly describes the brilliant spectrum produced by burning cyanogen with oxygen. Dibbits arrived at the same conclusion. In answer to the objection raised that carbon in the state of vapour could not exist in the flame of a candle or that of a Bunsen burner, Dibbits argues that carbon exists in combination with hydrogen in the gas which is burnt, and after the combustion that carbon is in combination with oxygen, and during the chemical change it may have existed as uncombined carbon ; a flame of carbonic oxide does not show the spectrum because the carbon was already combined with oxygen. In the case of cyanogen the carbon is at first combined with
* Astrophysical Journal,
vi. p.
65 (1897).
Ibid., xiv. p.
322 (1901).
62
An
Introduction
to
the
nitrogen,
and after the combustion it is combined with oxygen, and This spectrum is obtainable not so the same explanation appHes. only from flames but also by means of the spark in different gases containing carbon, either at atmospheric pressure or under reduced pressure in vacuum tubes. Thus it is given by the spark discharge in cyanogen, carbonic oxide, naphthalin, carbon disulphide, carbon tetrachloride, olefiant-gas, coal-gas, benzene, chloroform, aniline, These substances have ethylene dibromide, ethyl bromide, etc. only carbon in common, and unless the experiments are vitiated by impurities, the conclusion is irresistible that the spectrum in question In is due to carbon itself, and not to any compound of carbon.
o
1875 a paper by Angstrom and Thale'n was published, in which they They maintained that the spectrum was due to a hydrocarbon. seem to have been led to this conclusion by the resemblance of the spectrum to the spectra of metallic oxides. Their theory was adopted and maintained by Professors Liveing and Dewar, who observed the electric arc between carbon poles in various gases, such as air, hydrogen, nitrogen, chlorine, carbonic oxide, nitric oxide, and ammonia. They found the Swan spectrum well seen in the arc in hydrogen, but less strong when the arc was taken in chlorine or They nitrogen, but it seemed to he present whatever the atinosphar. account for this by the difficulty of completely removing all traces of hydrogen from the carbon electrodes. Lockyer, observing the electric discharge in the vapour of carbon tetra-chloride, found that it gave the Swan spectrum brightly, although there was no sign of the hydrogen lines. Liveing and Dewar found that the Swan spectrum was always to be seen in the spark between electrodes close together in the saturated vapour freed from moisture and air. Professors Liveing and Dewar, later on, and in consequence of their own further experiments, abandoned the view that the Swan spectrum depended upon the presence of hydrogen. The temperature of the arc is supposed to be about 3500 C, and Moissan has shown that carbon is distinctly volatile at that temperature. The flame of cyanogen in oxygen is the hottest flame known (except perhaps the flame of cyanogen in nitric oxide), and is far above 3500", and Liveing and Dewar admit that carbon vapour may exist in a cyanogen flame. Those who attribute the Swan spectrum to carbon, after having had to contend against the hypothesis that it is due to a compound with hydrogen an hypothesis abandoned by the very chemists who supported it so strongly now find themselves called upon to contend against a new theory advanced by Professor Smithells (on the suggestion of the late Sir G. G. Stokes), that the spectrum under discussion is due to a compound of carbon and oxygen. Professor Smithells has, with much skill, constructed a defence of his position, based for the most part upon a study of the phenomena of flames. He seeks to set aside the natural conclusion to be drawn from the repeated observations of the Swan spectrum in carbon compounds not containing oxygen, by the suggestion that it is impossible to free
...
Bana-spectra.
these
Spectra
of Compounds.
163
substances from impurities, water, etc., containing oxygen. " distrustful of conclusions based upon the supposed purity of the substances submitted to examination in highly exhausted Pllicker tubes. The difificulty of removing films of air or moisture from glass, the occlusion of gases by electrodes, the oxidizing character of glass
He
is
itself, all
form well-recognized
difficulties in
chemical
all
difficulties quite apart from the purely obtaining pure materials," and in this distrust
It is
true that
sometimes
in
vacuum
tubes the brightest part of the spectrum may be due to some substance present in very small quantity, as, for example, mercury in any tube exhausted by means of a mercury pump, but it would seem that the circumstances are altogether different when the spark is taken in gases at ordinary pressure, and still more so in the spark taken
in liquids.
To assume
to
is
always due
the
substance present in
the smallest quantity seems a rather the number of carbon compounds which do
not contain oxygen (or which ought not to contain oxygen), and which still give the Swan spectrum more brightly than anything else, is so large, that the conclusion naturally to be drawn from this accumulation of evidence will seem to most people to be that first drawn by
Attfield.
recent unpublished observations by the author on the spark seem to him to confirm the conclusion. Most of these liquids offer so much resistance to the passage of the spark that it is only possible to employ a very minute spark ; but when platinum electrodes, only a fraction of a millimetre apart, are used, there is no difficulty in obtaining a spectrum, the discharge being attended by rapid decomposition with separation (in most cases) of carbon, so that there is sometimes only a very brief time in which the spectrum
in hquids
Some
can be seen.
rapidly tion of
Thus the spark in chloroform (the purest that could be obtained) decomposes it with separation of solid carbon and evolutorrents
acid. The carbon so rapidly was found better to arrange a continuous flow of fresh chloroform, which carried the carbon with it. The electrodes were heated red-hot inside the liquid before the observation was made, but although the experiment was made many times, nothing but a brilliant Swan spectrum was ever observed. The same spectrum was obtained from the spark in the following benzene, liquids, in each case as pure as could be obtained In these naphthalene, anihne, ethylene dibromide, ethyl bromide. cases we see the decomposition of the liquid into carbon and something else going on all the time, and it seems impossible to resist the conclusion that the spectrum which they all give is that of the only element common to all, that is, carbon. By the Swan spectrum in these experiments is meant the spectrum shown in Fig. 16 (9), but in certain cases, more particularly when cyanogen is used, or when the arc is taken in air or nitrogen.
of
hydrochloric
it
164
An
Introduction
to
the
bands appear which are generally admitted to be due These are in the visible of carbon and nitrogen. spectrum a band of seven edges from 4600 to 4500, which we will call f, a band of six edges from 4216 to 4158, which we will call 6, a band of four edges from 3884 to 3855, which we will call and in the ultra-violet other bands observable in photographs at 3590-5, 3361. An interesting experiment which shows the dependence of some of these bands on the presence of nitrogen may be made by taking a This yields the Swan very short spark in liquid absolute alcohol. are not seen. On adding spectrum brilliantly, but the groups and a little strong ammonia the group B comes out very bright and also
certain other
to
compound
i,
t,
t,
less brilliantly.
There are two other spectra given by carbon compounds under One of these is generally admitted to be the spectrum of an oxide of carbon. It is given by a vacuum tube containing carbon monoxide, carbon dioxide, olefiant-gas, coal-gas, or It benzene, if no special precautions are taken to exclude oxygen. consists of about six bands at about 6060, 5610, 5198, 4837, 4509, and 4395. A vacuum tube filled with purified benzene, and containing pieces of bright metallic sodium, showed this spectrum at first, but after heating the sodium the spectrum had completely disappeared, and the tube showed the spectrum of hydrogen. On introducing a little ox)gen into the tube, by heating a fragment of potassium chlorate wrapped up in platinum foil which had been previously
certain circumstances.
comes
out.
third spectrum is produced by the spark with jar in carbon monoxide or carbon dioxide at the ordinary pressure. It is not unlike the spectrum of air shown in Fig. 17 (9), but the lines are quite
The
The chief features of the spectrum are a double red line close to the red hydrogen line, a double green line at 5152 and 5145,
different. line in the violet at 4267, besides some lines in the ultra-violet. This spectrum is, no doubt, the line spectrum of carbon. Since the Swan spectrum and the spectrum of cyanogen have been observed in comets and in some stars, the question of their chemical parentage becomes one of cosmical importance.
CHAPTER
XIII.
''
An important achievement of recent years is the taking of photographs of the Sun's disc at any time by means of the monochromatic light of any given substance, ^._f.. of calcium or of hydrogen. This is accomplished by means of the spectroheliograph. This instrument, the invention of Professor Hale of the Yerkes Observatory, was first used at the Kenwood Observatory in 1891, with a 12-inch equatorial refractor, but a much larger instrument has been made for the great 40-inch telescope of the Yerkes Observatory, which gives an image of the Sun seven inches in diameter, and with this instrument Professor Hale now photographs the Sun, with all its spots and faculse and surrounded by all its prominences, in a few minutes at any time, whereas formerly we had to wait for the rare occurrence of a total eclipse in order to see or to study the prominences. Other instruments of the kind are those of M. Deslandres at the Observatory of Meudon, of Dr. Kempf at the Potsdam Observatory, and at the South Kensington Observatory under the charge of Sir Norman Lockyer, and other important observatories will shortly be supplied with similar instruments. The principle of the method may be explained thus Across the image of the Sun formed by the telescope a slit is made to travel at a uniform rate by means of an electric motor, and an image of the light passing through this slit is formed by a spectroscope and thrown on a photographic plate, in front of which a second slit is made to travel at the same rate as the first, being driven by the same motor. If we suppose the first slit to be illuminated by monochromatic light, then in passing through the spectroscope it would suffer no deviation, we should simply get an image of the first slit on the plate ; and if the second slit is adjusted in the right position, we should get a photograph of the Sun's disc taken in successive slices. Of course, the first image is not monochromatic, but the spectroscope sends all the light not of the exact wave-length for which the instrument is adjusted elsewhere than through the second slit, and it is not admitted to the plate. If the adjustment has been made for the or line, we obtain a photograph of the incandescent calcium vapour forming part of the Sun's surface, and so for any other substance or line which is selected.
:
66
All Introduction
to
the
and K The ordinary visual spectrum of the faculte shows the bands reversed by a thin bright hne running down the centre of each. This indicates the presence at a high level of calcium vapour at a
higher temperature than that of the reversing layer below. Fig. 124 shows the spectroheliograph of the Yerkes telescope, and Fig. 125 is a copy of a photograph of a sun-spot made by means of it. In Fig. 121 we have a photograph showing the bands and widened lines in the spectrum of a sun-spot ; the most conspicuous group in Fig. i is the "little /'" group, as may be seen by a comparison of this photograph with Fig. 51.
l-'iu.
to 40-incli
Yerkes refractor.
In February, 1894, a most remarkable disturbance of the usual conditions of the Sun's reversing layer was recorded accidentally in a series of photographs reproduced in Fig. 126, in which the broad shading of the and lines usually attributed to calcium actually disappeared for a short time. In photograph No. i the dark streak is the spectrum of a spot, in which the and lines may be seen to be reversed on the spot-band. few moments later the spectrum
had changed
to that shown in No. 2, the broad shading of and has almost disappeared, the spot-streak is very faint in their neighbourhood, there are two very conspicuous bright lines at 3884'64 and 3896'2i, and there are other changes. After a very short time the spectrum had returned to its normal condition, as shown in
167
Such an extraordinary change involves an area of the Sun's 4. surface round the spot of a diameter of one-eighth of the Sun's diameter, and there does not appear to be any record of any similar
occurrence.
No.
Light is now known to be closely connected with electricity and magnetism. The first experimental observation of such a connection was made by Faraday so long ago as 1845, when he showed that a cube of a certain heavy glass (lead borate) when placed
Fig. 125.
Sun-spot of March
6,
1900.
between the poles of a powerful electro-magnet and examined by light sent through it, rotated the plane of polarization of the light when the magnet was excited, the direction of the rotation being the same as the direction of the electric current which excited the magnetism." According to the electro-magnetic theory of light of Clerk Maxwell, put forward in 1867, the waves of light are not due simply to mechanical vibrations of the ether, but depend upon electrical and magnetic vibrations at right angles to each other, and at right angles
means of polarized
Faraday's experi-
polarization in a transparent
* It is worthy of record that Faraday thought it possible that the bright lines of a spectrum might be affected by a magnetic field, and in the last experimental research that he made he tried the effect on the sodium lines in a flame placed between the poles of an electro-magnet, but without observing any change.
68
A?i Introduction
to
the
Study of
Spectriiiii
Analysis.
diamagnetic medium implies the molecular rotation of the ether. Plane polarized light may be resolved into two components of
circularly polarized light, one rotating with, the other against the ether. If the diamagnetic medium absorbs only one wave-length,
thus producing a single dark line, then when the magnetic field is excited the absorption of the one component will be of a wavelength greater, and of the other component of a wave-length less, than
if
split
the magnetic field did not exist. The dark line thus becomes up into two. This is known as the "^^Zeeman effect," as first observed by Professor Zeeman in 1896. To observe these effects very
fields
powerful magnetic
great dispersion.
are necessary,
efficient
The most
the lines will be different according as they are seen in the direction of the lines of force, or at right angles to them. Since the effect of the magnetic field is an acceleration in the one direction, and a retardation in the other, round the lines of force, when viewed in the direction of the lines of force the result will be that the two com-
ponents are circularly polarized in opposite directions, and if they then pass through a quarter-wave plate to convert them into plane polarized light, and then through a Nicol's prism, the spectroscope will show a displacement in the issuing light on turning on the
current.
the
same
The direction of rotation of the accelerated component is as that of the magnetizing current, because in the magnetic
the molecules are rotated round the lines of force in this If light consists in the motion of electrically charged particles about a position of equilibrium, their motion can be resolved into three components, one in the direction of the Hnes of force, and
field
direction.
at right angles to the lines of force, and these last can be resolved either into right-handed or left-handed rotations. The magnetic field acts only on these last, and according to the direction of their motion and the nature of the charge, whether positive or negative, they are pressed towards or drawn from the position of equilibrium, and there results either an increase or decrease in oscillation frequency. The observation that the accelerated component has the same direction of rotation as the magnetizing current leads to the conclusion that the particles are negatively charged. At right angles to the lines of force the edges of the bands are plane polarized, the plane of polarization being parallel to the lines of force, since the vibrations are at right angles to the lines of force. Zeeman concludes that, seen at right angles to the lines of force, a line in the spectrum should be resolved into a triplet, the central component of the triplet corresponding to vibrations in the same direction as the lines of force, while the two outer components correspond to the vibrations at right angles to the lines of force, and he actually observed this to happen with the sodium lines, in 1896. Zeeman, in 1898, observed that the splitting up of the different lines in the same spectrum is very different in the same magnetic field. In the zinc spectrum the components of the lines marked t are
two others
169
Tlie "
Zeeinan Effect."
171
widely separated, whilst those of the lines marked * are hardly perceptibly divided. Upon further examination it appeared that the phenomena were
more complicated than at first supposed. Michelson showed that the middle component of the triplet into which any spectral line becomes resolved in a magnetic field was itself a symmetrical triplet, that the relative intensity of the components varies for different substances and for different lines of the same substance, so that the
appearance may be that of a single
line, a double line, or a triplet the outer lines are unsymmetrical, but symmetrically placed with respect to the central line ; their intensities vary so that the outer groups may present the appearance of single, double, or multiple lines. In the sodium spectrum the least refrangible of the yellow lines is split up into four components, and the most refrangible line into six components ; in the first case the central line of the normal triplet becomes doubled, and in the second case each line of the triplet
becomes doubled. Still more complicated effects have been observed in the spectra of mercury, zinc, cadmium, magnesium, etc. It is a most interesting fact that the homologous lines of related elements whose wave-lengths can be calculated from the formulte of Kayser and Runge seem, generally, to be affected in the same way by the magnetic field, and that in the same spectrum the lines of the first subordinate series are
affected differently from those of the second subordinate series.
different behaviour of
The
127,
some
lines of
mercury
is
shown
in Fig.
reproduced by the kind permission of Professors Runge and Paschen. The line 546j'o is seen to be resolved into nine lines, of which the three in the centre are polarized in one plane, and the outer six are polarized in a plane at right angles to the first three. This is the first line of the triplet in the second subordinate series for which = 3. The second line of this triplet, 4358'6, is resolved into four lines ; the line 4046'8 is not one of the series lines, it is resolved simply into three lines. Preston has observed that the separation of the components bears the same ratio to the square of the wave-length in the corresponding series-lines in the same element, and that this remains the same for corresponding lines in the homologous spectra of allied elements. The pairs of lines in copper and silver show the same behaviour as the two sodium D lines, except that it is the line of less wavelength which is split up into four like Di, and the other line of the pair which is spht up into six like Do. The wave-lengths of the components of Di are sSgS'or, 5897'09, 5895'26, and 5894'28 in a certain field, 'and the components of Do in
same field, 5892-55, 5891-54, 5890-66, 5889-72, 588876, and 5887-91. a In the case of the mercury line, 5461, shown in Fig. 127 field of about 24600 C.G.S. units, the components had the wavelengths 5461-61, 5461-451 546i'29. 546i"i3> 5460-97, 546o-8r, There appears to be no splitting up 5460-65, 5460-48, and 5460-32.
the
The
''
Zeeninn
Effect."
173
of the bands of mercury or of those of nitrogen or of cyanogen in the magnetic field. Professor Lorentz has given a theory which goes far towards explaining the Zeeman effect. He assumes that in the luminous gas minute electrically charged particles of definite mass and charge are present, and that the vibrations of the light emitted are determined by the vibrations of these particles. The assumption that every molecule of a source of light contains one such electron vibrating
by a force proportional to the displacement, leads to the conclusion that a single line will be tripled by the action of a magnetic field if viewed at right angles to the lines of force, and will be doubled if viewed along the lines of force, and further, that the components will be polarized in the manner observed. Professor Zeeman gives the following account of the theory " Imagine an observer looking at a flame placed in a magnetic field
position
:
towards or from him. Let us suppose that the observer could see the actual ions as they are revolving then the following will be remarked There are some ions moving in circles and hence emitting circularly polarized light if the motion is round in the positive direction, the period will be longer than with no magnetic field; if in the negative direction, shorter. There will also be ions seemingly stationary and really moving parallel to the lines of forcq with unaltered period. In the third place there are ions which seem to move in slowly rotating
in a direction such that the lines of force run
;
:
elliptical orbits."
theory of Lorentz does not account for the since been observedj and in a later paper Lorentz modifies the theory by substituting for the single electron a system of electrically charged particles having more than three degrees of freedom, viz. as many as the components observed in the magnetic field.
But
this
original
CHAPTER
XIV.
guished from
resolving power of a spectroscope must be clearly distinits dispersion. The latter is a measure of the angular separation produced between two given monochromatic rays, while the former is a measure of the number of monochromatic rays which it is possible to distinguish as separate between two given monochromatic rays, and is the really important criterion of the power of the instrument. For it must be remembered that the
The
image produced by a telescope the termination of every spectroscope is not the indefinitely thin line given by the construction of geometrical optics, but one of appreciable breadth, shading oft" to The smaller the aperture of the telescope darkness on each side. (as limited either by the diameter of its own object glass or by the width of the beam of light from the prism or grating system) the If then the wave-lengths of two rays broader do the lines become. differ by so little that these lines overlap with a particular aperture, an increase of aperture alone, provided it be sufficiently great, is
The resolving power of a grating can then be indefinitely increased by ruling lines over a wider area without alteration in the number of lines to the millimetre, for the effective aperture being increased the spectral lines become finer and hence separable. The resolving power is in fact proportional to the length of ruling that is, to the travel of the grating plate during the cutting. Unfortunately the difficulties of cutting accurate gratings increase very rapidly as the length of ruling increases, so that gratings with 5 or 6-inch ruled surfaces are the biggest that can be obtained, and even these are very difficult to get.
There
resolving
is
power
another way of getting higher resolving power. The is given by the product m ;/, where in is the order
175
of spectrum observed and n the number of elements of which the the number of divisions, that is, in the case grating is composed
of ruled gratings. With the ordinary form of grating the intensity rapidly diminishes as the order is raised ; but if a grating could be constructed to throw most of the light in one of the high orders the desired object would
be
attained.
result
device which was first suggested by Michelson to obtain this was as follows Let a number of plates of silvered glass of equal thickness, /, be superposed, each projecting beyond the last by an equal amount in such a manner as to form a series of steps, as shown in Fig. i 28. If then a beam of " parallel " light be incident in the direction of the arrows, it will be reflected from the successive steps, and the reflected beam will consist of a number of beams reflected from the first, second, third, etc., plates, whose retardations behind that reflected
:
The
<
^~T
Fig. 128.
Reflection echelon.
from the first plate will be o, /, 2t, 3/, etc. They will, in fact, be in the requisite relation to produce, when brought to a focus by a lens, diffraction spectra of the light source. If t be taken equal to 10 mm. the orders of spectra observed
would be in the neighbourhood of the 34,000th, so that five plates would give a resolving power greater than that of a lo-inch grating
in the first order.
however, unnecessary to go into the analytical expressions it has not yet been found possible to manufacture it. The accuracy required in the parallelism of the plates in order for the grating to perform in accordance with the theoretical expressions of its capabilities would be such, for one of ten plates, that no differences in thickness should exist among the different plates so great as one-twentieth of a wave-length of light, that is, so great as oneThis, in itself, is not outside the limits of posmillionth of an inch. sible accuracy it is, in fact, reached in some of the echelons made by Mr. Hilger. But the plates must lie on each other in perfect contact no gap existing between even the extreme edges of successive plates of as much as the quantity stated above, and this is the difficulty which has not yet been surmounted. The form which has been manufactured with success, and which is now well known as the echelon grating, is Professor Michelson's
It
is,
176
A;/ Introduction
to
the
modification of his original idea, an account of which he pubUshed in the A sfrophysical Journal for June, 1898, and subsequently in the Journal dc Physique, 3rd series, tome viii., 1899. It is from these sources that the following analytical investigations are taken.
The arrangement, which is illustrated in Fig. 129, consists, like that described above, of plates of parallel glass of equal thickness superposed in step form but the glass is not silvered, and the beam of
:
<
1
^
<-
*
I
<
< <
<
I
-<-
^
1
,
p^^
Fig, 129.
Transmission echelon.
177
light
be incident
= width of the steps {a b, Fig. 130). = wave-length of the light under consideration, = refractive index of the glass for light of wave-length m = retardation in terms of X between rays emerging at
ju.
A,.
angle of diffraction
of successive elements.
Then
mk =
ixt fxt
=
Since the values of
ad
t
cos
-\-
sm
B
this
may be
written
(i)
mX =
(im
i) t
6s
constant,
terms
de,
d\
or, substituting for
= '-(m-t
s\
dX
(/j.
its
appro.ximate value
'
i)-
X
,
.
dOi
^^>
dX
^ TXL
sX
tf,
^'
'2'H
^dXi
'
'
(.dnpcrswH)
value of the co-efficient b can be found for any wave-length from the optical constants for the glass used (it is between o'S and 1*0 for most glasses), and the equation (2) will then give the angular dispersion dOi between two homogeneous radiations whose wavelengths differ by a small amount dX. Differentiating (i) with respect to m, X constant, and rearranging
the terms
The
dO
_X
s
dm ~
by
dii,
Putting dm then
i,
and representing
(3)
'-~~s
'
'
'
{separation of spectra)
Equation
orders of spectra.
178
Ah
dO,, represent the angular limit of resolution of the echelon, the angular separation of two spectral lines when they are just distinguishable as separate in the eyepiece of the telescope. By the well-known rule (see Lord Rayleigh on Wave Theory, " Encyclopedia Britannica ")
Let
i.e.
(,4;
tts
}l
Then from
(2)
de,
bt
dX,"
or, substituting
-;
s\
for dO^
(4(i')
of resolution)
corre-
dX. in the
homogeneous
and the
is,
ratio
will
-ir-
is
This
it
be seen,
proportional to the total thickness of glass traversed, and for a given wave-length is independent of the thickness of plates or width of step. The brilliancy of any line seen with the echelon depends not only on the intrinsic intensity of the radiation, but also on the angle of
diffraction, 0.
The
co-efificient
which depends on
is
0,
be called the
intensity of illumination,
sm
(S)
ir-b
for Dr.
Thickness of plates,
Width of
steps,
/= =
about 10
i
mm.
mm.
179
/i
^F
/xi
= = = =
Whence
-J^
for
wave-length
5890 (D)
-7-19 x
lo"",
Number
(It is to
of elements, n
21.
be noted that when a clear aperture of i mm. is left beyond the width of the largest plate, the number of apertures operative in the formation of the spectrum is one more than the number of plates. It is this number of apertures which is denoted
by
.)
From
(2), p.
177
dO^
bt
The
Dispersion.
^=-^-,where^=(/.-i)-A^
,
djx.
Substituting for u, A,
and^,
dk
b
D,
= =
0-575 0-617
0-042
Hence
-r-*
=
=
0-00105 i"
o'l, dO^
''^^
region of D.
Putting dk
0-000105 radian,
differ in
The Separation
From
noted by
(3), p.
dO-i,
= D
Hence
for
dO.,
orders are, then, very close together, and an is necessary for preliminary analysis of the light ; only a very restricted portion of the spectrum under observation being thrown on the slit of the echelon spectroscope.
The
successive
auxiliary spectroscope
l8o
/I
Introduction
to
the
of Resolutiox.
'-.
Hence
for
D
dO,.
=
'-
(for
From
(4(7), p.
178, the difference of wave-length necessary between in order to distinguish them as separate,
= = ^ but
,
X-
(for
^
7 D) '
o
5890"
r,
0-617
2[
::
10
Angstrom
units
=
It
single ray
must be noted that one can determine the displcueinent of a more readily than one can separate the components of a
In fact, a shift of a line of as little as one-tenth the angular limit of resolution can be detected. (See " Le Phe'nomene de Zeeman." A. Cotton, p. 15.)
close doublet.
Increase in thickness of plates increases the dispersion, and thereamount of detail distinguishable. The separation of spectra
remains the same. Increase of ii<idth of step decreases the separation of successive spectra but the angular limit of resolution is reduced also, the amount of detail distinguishable remaining the same. Increase of number of plates does not alter dispersion or separation of spectra but it reduces the angular limit of resolution, the amount of detail distinguishable being thereby increased. In addition to this
;
;
the
amount of
light is increased.
will
be seen that
when
-.
Now
;
is
the angular
separation of successive orders of the same line hence if any spectral line be formed at 6 0, the two orders next it on each side are absent, and practically all the light is concentrated in the one line.
Grating.
I8l i8
This fact should be well noted, as by a slight rotation of the echelon it is found that any desired line can be got into this position, and the property above mentioned is one of great importance in the practical manipulation of the echelon.
Fig. 131.
Intensity curve.
etc.
Mounting,
Mr. Hilger, who has made echelons for many of the principal Universities of Europe, recommends the following dimensions as those which have been found most generally satisfactory, and which he always 'supplies when no contrary instructions are given in
ordering
:
Thickness of plates
10
i
mm.
Width of step
wrwvm.
Fig. 132.
varies
somewhat according
to the
number
82
A>/ Introditclion
to
the
The
largest
on an optically worked metal plate in which the aperture is cut, and the plates are clamped by steel rods with millhead nuts. The plates are cleaned with the most scrupulous care, and when clamped in the above manner they make optical contact
over the greater part of their surfaces. By this means the great loss of hght noticed by Professor Michelson (who did not clamp the plates of his echelons) is obviated to a surprising degree, and the more recently made echelons are remarkable for their brilliancy ; while experience shows that, contrary to expectation, judicious clamping does not cause the slightest loss of The smaller sizes of echelon can be adapted to almost definition. any form of spectroscope, but owing to its peculiarity of being always used at direct vision, a special form becomes more convenient. Such an echelon spectroscope, designed for a twenty-plate or larger echelon, is shown in Fig. 133 in conjunction with a special
Fin. 133.
form of auxiliary spectroscope (known as the " Constant Deviation Spectroscope '') which presents many advantages for the purpose, the chief of which is that neither collimator nor telescope is ever moved. Arrangements are provided by means of rods and bevel gear for the various adjustments that have to be made, such as the passing through the spectrum by rotation of the prism of the auxiliary spectroscope, the opening and shutting of the slit, the rotation of the echelon to get the line under observation into any desired position, and the moving of the echelon out of the field
of view.
little
up the apparatus in the first instance, care being specially taken that the whole of the echelon is operative. To do this one should proceed
as follows
:
183
Set the echelon and auxiliary spectroscopes up end to end as shown in Fig. 133, with the axis of each horizontal and at exactly The eyepiece should, of course, the same height from the table. be removed from the auxiliary spectroscope, and the distance from the end of its telescope tube to the focus of the telescope object glass being carefully measured beforehand, the auxiliary spectroscope should be put at such a distance from the slit of the echelon spectroscope as to bring the spectral lines to a focus on the slit. Set the telescope of the auxiliary and the collimator of the echelon Now spectroscope in one line as nearly as can be judged by the eye. put a sodium flame in front of the auxiliary slit, and turn the prism of the auxiliary spectroscope until D comes on to the " echelon " slit. Then on looking through the echelon spectroscope with the eye-
piece removed, the object glasses should appear equally illuminated except, perhaps, two slight black strips right and left, which, When this is the case if present, should be perfectly symmetrical. (and the setting is not right until it is so) the apparatus is ready. If now the slit of the auxiliary spectroscope be closed to a convenient width and that of the echelon spectroscope be opened wide, the spectrum of any light placed in front of the slit of the auxiliary spectroscope will be seen in the eyepiece of the viewing telescope. Any particular line that it is desired to examine must then be got on to the fixed jaw of the slit, the slit closed to exclude all other Supposing the radiation to be lines, and the echelon introduced. homogeneous within the limit of resolution of the echelon, the following possible conditions may be observed in the eyepiece 1. There may be one bright single line visible, together with a series of very faint lines gradually diminishing in intensity as the This is the condition distance from the central line is increased. in which the two neighbouring orders to right and left are absent
all over,
:
There
may be two
lines
of
equal
line,
intensity,
which
are
very faint
field.
3. There may be any number of conditions intermediate between these two, with two lines visible, but of unequal brilliancy, and the The series of lines to right and series of very faint lines as before. left will not be found to interfere with the use of the instrument in there are two points right and left which are the slightest degree always dark and which enclose a space equal to twice the separation of successive orders (see p. 181). It is on this region that the attention
should be concentrated. By a slight rotation of the echelon about a vertical axis any of indeed, one can the conditions above mentioned can be attained pass a large number of orders through the field one after the other. There is one position of the echelon, however, at which, during
motion
in
much
is
same way
when
the prism
the rotated
84
An
Introduction
to
the
Study of
deviation.
light,
Sfiectrian Analysis.
minimum
This
is
when
the echelon
and
this position, or
near
it,
Such are, the determinastructure or of variation of spectral lines. tion of the effect of pressure on the nature of the radiations emitted by glowing gases, etc., the effect of motion of the light source relative
to the spectroscope,
and above
"
to
all
Zeeman
The
one or two cases. difference between the velocities of two opposite points on
is
this
produces a
As
in the grating
investigated above a shift of 0-003 A.U. would be perceptible, it seems probable that the echelon may form a valuable tool in the
hands of astronomers.
The displacements towards the red end of the spectrum, due to a pressure of eleven atmospheres in various lines, have been given by Humphreys (from experiments made in Rowland's laboratory with the 6-inch reflection grating ; see the Astrophysical Jourtial,
October, 1897, p
o
169).
The
greatest
among
A.U.
In conclusion it may be remarked that there is no difficulty in constructing echelons with plates considerably thicker than 10 mm. (Professor Michelson, in fact, made them with 20 mm. plates), and thus increasing the resolving power, should this be desired.
CATALOGUE OF SPECTRA.
Table of Standard Wave-lengths
(in air at 20
1893,
xii.
321.
Reduction
Wave-lengths.
and
character.
Element.
Kind of standard.
In arc In sun
OS
2-09 2-08
3-5
[A-
[Ai.]
.]{o (0
'10
U
14 14 14 12
]-2
8 8
7714' 680 7699' 374 7671' 994 7670- 993 7666' 239 766o' 265 7C00' 778
76.')9'
2-07
658
2-OG
3-6
[A*] [AtJ
O n
10
754o" 921 7511' 286 7495' 351 7462' 609 7446' 038 7409' 554 7389' 696 7331' 206
2-04 2-03
12958-8 12984-6 13030-9 13032-6 13040-7 13042-4 13050-0 13051-9 13105-1 13107'4 13111-5 13113-7 13117-6 13164-6 13249-2
13.309-7
2'o2
2'oi 2-00 1-99 1-98
wvV wv y
"WV?
wv ? wvV
"WV?
1^"'
C
I
7321 056 7318' 818 7304' 475 7.300' 050 7290' 714 7287- 689
S
'.
8 8
i
III.
JII.
7273' 2.56 7270' 205 7265' 833 7264' 851 7247' 46 7243' 904 7240' 972 7233' 171 7232' 509
72-27
13338-0 13396-5 13426-4 13492-5 13528-7 13636-6 13655-5 13659-7 1368G-5 13694-8 137T2-4 13718-1 13745-3 13751-1
1.3759-3
1-96
.
i
765
8
(i
1:1.
wv?
* Single line
Note.
Unless
of about 20 C.
pressure.
wave-lengths are upon Rowland's scale All oscillation frequencies are in vacuo.
86
A?i Introduction
to
the
contintu-d.
Element.
wv
\S \v \s
V V V
;
':
':
wv V wv
wv
wv
\\' \'
? V
(
I
[Bu.]
('
[Bn]
[B,o]
10 (0
(
fo
10
Ni
'
'.
[IM
[Br]
[B,:]
10 (O
10 fO lo
[1>-,]
JO
lo (o 10
[B,]
Catalogue of Spectra.
OF Standard Wave-lengths (rtf//;&v/.
187
'J'AiiLE
88
Au
Introduction
to
the
Catalogue of Spectra.
Table of Standard Wave-lengths continued.
189
Intensity
and
character.
We
Kind of
standard.
Element.
In
arc.
In sun.
n.
II.
I.
I. J.
J.
I. 1. I. J.
I.
J.
I.
\. I.
I.
]. I.
I.
I.
I.
15r
I. I. ].
I.
II.
15r
I.
I.
II.
I.
M.
IV.
lOr
I.
II.
I.
T. I. I.
I.
I.
I. I.
IV.
10 10
I. I.
I.
IQO
An
Inti'odtiction to the
continued.
Reduction
Intensity
and
character.
Weight.
Wave-lcngtlis.
Element.
In
arc.
Kind
In
sun.
vacuo.
of
standard.
Fe Fc
Fe Ft
^vv
4
6
Fc Fe Fe
Si
Fe Fc
wv
Fe
?
wv
1
Fe Fe Fe
wv
Fc.-
[Di]N Ni [D,]\.
.
Fe
[Ua]ll
Fe Fe Ca
Ba
Ni Fe Fe Fe Ni Fc
V
Fe
(Jr
'
I
/
.
Cr Cr
Cu Fe
Si.
Cu
. . .
.
Fe
'I
Ni Fe Fe Fe Fe
J. .
Catalogue of Spectra,
Table of Standard Waa'e-lengths
191
continued.
192
A// lutrodnction
to
tlie
continued.
Reduction
,
and
character.
Weight.
Wave-lengths.
Element.
In
arc.
Kind of standard.
In
sun.
In
arc.
In arc
(<x).
Ni
II.
III.
11.
III.
Fc Ff Fe
*
.
]. I. I.
Vm
3
4 A
nil-
Va
F.'
Cr Fc Fc Fe Fe Fc Fc Fc-Cr Ni Fc Fc
.
. . . . . . . .
I.
I.
I.
I. I.
I.
}IV.
J I.
1.
IV.
Fc iyu)
Fc-Xi T1,J
C:i
.
111.
III.
IV.
M.
JI. Ill
11.
Fc Fc Co
Fe Fe Cr Cr Fc Fc Fc
Cc.
y
.
I.
>.
111.
[1I7J]
?
.
III.
HI.
I.
I.
,
I. I. 1.
I.
Cr
FcJ
Fc
Cii
F:
H
)
}
I.
II.
I.
M.
*
Ill
I.
jH. Ill
1.
Fe
Catalogue of Spectra.
Table ok Standard Wave-lengths continued.
193
Intensity
and
character.
Weight.
Wave-lengths.
Kind
of
j
standard.
In
arc.
In
sun.
[Fc]
4r
Ca Ca
Cr
j
. .
8
(i
Ca
4r G
i}
Cr V Caj-
2
6
I
Ca
Fc
Fc
Fc Fo Fe Fe
Fe Fe Fe
Ti
lUr
8r
Fe Cr
>
FeJ
.
Fc
Ti
Ca
'I
Ti
'J'i
[bI]M.^'
.
[b.]Mg Fe Fe )
.
Fe Fe
j
]
Fe
C*
Fc Fe? Ni
Ti
?
Co
J
Mai
Fe
Ni
194
^'J"
Introduction
to
the
continued.
Intensity
and
character.
Kind of standard.
In
arc.
NiJ
Fe Fcl
.
Fel
Fc '] Fc Co Fel
.
Nil
Ni
Fe
.
.
Fc Fe (Cu) Fe Fe
. .
Cd
Fc Fe
Ti
Fc Fe
fa
Xil
Ti J Ti Ti (Ni) Ti
.
Mg
b'd
.
Nebula Fel
Tij-
Fe Fe
I'b
Ti-La
lllr
.
1 Ol-
Fe
Ti
.
I
io
^}
Fe\
i
i
.
Fc
Ti
Nebula Fc
.
Fc
Catalogue of Spectra.
Table of Standard Wave-lenuths continued.
19s
ig6
All Introduction
to
the
continmd.
197
igS
An
Introduction
to
tlic
Catalogue of Spectra.
Table of Standard Wave-lengths continued.
199
200
A// IntrodtictioH
to
the
continued.
201
continued.
202
A)/ Introduction
to
the
continued.
and
character.
Element.
Kind
of
standard.
Fe
Ti Cn-Ti
1
n.
].
3,
(i
Mn
Fe* Fet
'
2
111
1(1
:}
M. IV. M. IV.
'}
r]
Nal
IV.
I.
Ka/
Mn-Di
^
lOr 15r 3, 2
4. 7
AI.
I
M.
I-
'"-Ti
Fe Ti
Cii
5
3(Jr
II)
Va
"I MnJ
Cu
.
4 4Ur
Fel
'
Ti Ti
lOr G
Ti}
Fe Ti
''
.
\
?
I
.
Fe Fe Fe
Ti
Fe
Ti
lUr
Ni
Cry La? F
;
I\ln
Ca
Fe Fe Co Ni Zr
A'a
4
lOr
Fe
Cr?
Ni
Catalogue of Spectra.
203
Table of Standard
\VAVE-'ve.viGTa%co?iiinueci.
204
A// Introduction
to
the
Table
(jf
Intensity
Catalogue of Spectra.
Talle of Standard Wave-lengths continued.
205
Inteiibity
2o6
A/i Introduction
to
the
Study of
Spcctniiii Analysis.
continued.
and
character.
Weight.
Kind of
standard.
In sun.
In
arc.
I
In
sun.
]
arc.
Fc
Si
M.
FeV
Si Si Si
Fe Fc
Fe
Fc Ca
Fc Fc Fe Fc Al k\ Fe Fe Fc
'.
Ba Ba
FeV Ca
Sr
21
)r
Jlir
1
20r lOr
A
)i
k\ Al
Si Si Si Si Si
Sr Sr
Catalogue of Spectra.
207
AIR.
Spark-spectrum or Elementary Line-spectrum.
Wave-length.
3o8
A/! Introduction
to
the
ALUMINIUM.
Intensity
:iiid
character.
Wave-length.
Spark.
TO.".:-'.!
Catalogue of Spectra.
209
ANTIMONY.
Wave-length.
210
An
Introdiiction to the
Catalogue of Spectra.
211
Argon continued.
Wave-length.
212
An
Introduction
to
the
Catalogue of Spectra.
213
BERYLLIUM.
Wave-length.
214
An
bitroduction
to
the
BORON.
Spark-spectrum.
BROMINE.
V.\cuuM Tube Spectrum.
Catalogue of Spectra.
'BROUlNE.cofiiin?ied.
215
Wave-length.
2i6
/I
Introduction
to
the
CAESIUM.
Intensity
:ir
Wave-length.
.S',M',)-9
86-27-7
* 6973-9
(2)
* 6723-6
(8)
* 6010-6
0)
* 6846-1
(2)
* 5664-0 * 5635-1
(13)
Catalogue of Spectra.
217
CALCIUM.
2i8
Ah
Introduction
to
the
CARBON.
Line-spectrum.
220
A// Introduction
to
the
Cerium continued.
Intensity
and character.
Wave-length.
""
Catalogue of Spectra.
221
222
An
Introduction
to
the
Chromium continued.
Catalogue of Spectra.
Cobalt
223
continued.
Wave-length.
24
Ati Introduction
to
the
C OB ALT
Intensity and characte
Wave-length.
Spark.
-
Catalogue of Spectra,
Copper
225
continued.
226
An
Introduction
to
the
CYANOGEN.
Catalogue of Spectra.
227
228
An
Catalogue of Spectra.
229
GOLD.
230
An
Introduction
to
the
HELIUM.
Catalogue of Spectra.
231
INDIUM.
232
A>i Introduction
to
the
Catalogue of Spectra.
Iridium continued.
233
234
^>i Introduction
to
tJic
Iron- -continued.
Wave-length.
Catalogue of Spectra.
Ikon
Wave-length.
235
continued.
236
An
Introduction
to
the
Iron
Wave-length.
continued.
Catalogue of Spectra.
Iron
Wave-length.
237
continued.
238
An
Introduction
to
the
lEON- -continued.
Spark.
Wave-length.
Wave4ength.
4158-9 4157-9
(1)
415C-9
(1)
4149-4 4147-7
(4)
4144-0 4143-6
(10)
10 10
4137-2
(3)
8
10
4134-8
(3)
4133-0 4132-2
(7)
8 10
6 (8)
4127-7
(9)
4122-6 4121-9
(2)
6
6 6
3998-2 3997-5
(8)
4120-3
(3)
3986-3
(1)
4118-7
(8)
10
3984-1
(1)
4114-6
(7)
3981-9
(4)
4109-9
(2)
3977-9
(11)
4107-6
(5)
3971-4
(3)
4104-2
(*)
3969-4
(17)
4100-9
(3)
3956-8
(6)
4098-3
(2)
4096-1
(17)
6 6 8
6
3948-9
(8)
3942-6
(1)
4080-0
(3)
3941-0
(4)
4078-4
(2)
6
8
3936-0
(S) 39-28-1
(3)
4076-7
(2)
4074-9
(4)
3
8r
G
3923-1
(2)
4071-9 4070-9
(1)
10
3920-4
(3)
4068-1
3917-3 3916-9
Catalogue of Spectra.
Iron
Wave-length.
239
continued.
240
An
Introdtiction to the
Iron
Wave-length.
co?ilinued.
Catalogue of Spectra.
lKi:>ii conlinued
241
Wave-length.
242
A^i Introduction
to
the
I RON
Wave-length.
Catalogue of Spectra.
I
24:
RON
continued.
Wave-length.
244
A"
InU'oduction
to
the
Iron
continued.
Wave-length.
246
An
Introduction
to
tJic
I Riix
co7ttilined.
Wave-length.
Catalogjic of Spectra.
247
Iron
continued.
Wave-length.
248
An
Introduction
to
the
LANTHANUM.
Catalogue of Spectra.
249
Lanthanum continued.
250
An
Introduction
to
the
Lkad continued.
Wave-length.
Catalogue of Spectra.
!5i
MAGNESIUM.
252
A// Introduction
to
the
MANGANESE.
Catalogue of Spectra.
253
Manganese continued.
Wave-ltngth.
2 54
A>i Introduction
to
the
M ERC UK Y
Catalogue of Spectra.
255
MOLYBDENUM.
256
All hitrodiiction
to
tJie
Study of
Spectruiit Analysis.
Molybdenum continued.
Wave-length.
Catalogue of Spectra.
257
Molybdenum continued.
Wave-length.
258
A>/ Iiitrodiictiou
to
the
Molybdenum continued.
Catalogue of Spectra.
Moi^VBDENVMconttfmeit.
j
259
Intensity
and character.
26o
A)/ Introduction
to
the
XeodiuyiMIUM
continued.
\\'a\'e-Iength.
Catalogue of Spectra.
261
NICKEL.
Wave-length.
262
A/i Introduction
to
the
Nickel
coittintied.
Wave-length.
Catalogue of Spectra.
Nioi3lVMconitnue<L
263
Wave-length.
264
All lutrodiictioii
to
the
NITROGEN.
Spark-spectrum or Elementary Line Spectrum.
266
^lu Introduction
to
the
OXYGEN.
Sl'ARK-SPECTRUM OR ELEMENTARY LINE SPECTRUM.
Catalogue of Spectra.
267
PALLADIUM.
268
Ah
Introduction
to
the
Catalogue of Spectra.
269
POTASSIUM.
The
lines
marked
PRASEODIDYMIUM.
2/0
A// Iiitroduction
to
the
Praseodidymium
continued.
Catalogue of Spectra.
FRASEODWYMlVMroniinued.
271
Wave-length.
272
An
Introduction
to
the
Catalogue of Spectra.
273
R H(JIJ UM
I
continued.
Wave-length.
274
-i Introduction
to
the
Rhodium contimied.
Catalogue of Spectra.
275
Rhodium continued.
Wave-length.
276
A// Introduction
to
the
RiiowUM
Catalogue of Spectra.
277
RUniDIUM.
Wave-length.
RUTHENIUM.
2/8
ybi Introduction
to tlie
RuTUENiUili:mtinued.
Wave-length.
Catalogue of Spectra.
279
RuTllEmvucontinued.
Wave-length.
28o
An
hilroiliiction
to
the
Catalogue of Spectra.
281
SELENIUM.
Wave-lenglh.
282
An
Iiitroditctiou to
the
Catalogue of Spectra.
283
SODIUM.
Wave-length.
284
All Introdiictioii
to
the
Study of
Spcctriini Analysis.
Strontium
continued.
\\'n\e-Iength.
Catalogue of Spectra.
285
TANTALUM.
Wave-ltngLh.
286
An
Introdiiction to the
TELLURIUM.
Intensity and character.
Catalogue of Spectra.
287
THALLIUM.
288
Au
conttJiued.
\\'ave-length.
Catalogue of Spectra.
TllORl UM
coHtimicd.
289
Wave-length.
290
All Iittroduction
to the
TIN.
\Va\e-Iength.
Catalogue of Spectra.
291
TITANIUM.
Wave-length.
292
A/i Introduction to
tlie
Titanium
Intensity
continued.
Intensity
and character.
Wave-length.
Spark.
and character.
Wave-length.
Arc.
Spark.
4623-2
(2)
4617-4
10s
Ss 8s 8s 8s
10
7
d88o-2
(2)
03)
4590-1
(7)
4870-3 4868-4
(1)
4572-2
()
20
15 8
9
485G-2
(2)
4563-9
(5)
4841-0
(3)
1655 '(J
(3)
J820-C
(V)
8n
10s
Is
4805-G
(1) 4790-ii (2)
15r
8
4792-6
(1)
8s
63
4778-4
(2)
4759-4
(1) 4758-:;
(I)
lOs
10s
8n
9 10 8
10
H)
4527'5
(3)
4742-9
(3) 4731-.S
8s
4523-0
(1)
4518-2
4512-9
(5)
4723-3 4722-8
(3)
4710-3
(a)
4501-4
(1)
15
4698-9
(2)
4496-3 4495-2
(3)
4693-8
(1)
4691-0
(S)
4489-2 4488-0
(2)
4682-1
(3)
4481-1
(2)
4675-3
(2)
4475-0 4471-4
(2)
4667-8
(5)
4656-6
(5)
4468-7 4466-0
(5)
15 5
10
4650-2
(1)
4457-6
(1)
4645-4
(2)
4455-5
(J)
4640-1 4639-8 J
14639-5
(7)
4453-9 \ 4463-5
f
(1)
4629-5
4451-1 4450-7
Catalogue of Spectra.
Titanium
293
continued.
Wave-length.
294
A)i Introduction
to the
Titanium continued.
Wave-length.
Catalogue of Spectra.
Titan U m continued.
i
1
295
2g6
A/t Introduction
to the
Catalogue of Spectra.
IV^GSTWicontimied.
297
298
All Introduction
to the
Tungsten
lon/iniied.
300
A7i Introduction to
ilic
Catalogue of Spectra.
'VA.SA.-Divil continued.
301
Intensity
and character.
Wave-length.
Arc.
(7)
4065 '2
(10)
4051-5 4051-1
(U)
4035-S
(13)
4023-5
(18)
40U6-9
(1)
4003-1
(5)
3997-3 3992-9
(3)
3990-7
(12)
3977-9
(1)
3973-8
(10)
3952-1
(32)
3916-6
(2)
3914-4
(79)
3847-5
(44)
3816-6
(23)
3795-1
(7)
3787-6
(10)
3778-8 3778-6
(9)
3773-1
(3)
3771-1
(11)
4
2
3760-4
(13)
3750-1
(2)
3746-0
(7)
4s
3736-2
(2)
3732-9
(6)
4s
3728'6 3723-5
(13)
3718-3
302
All Introduction
to the
Vanadium continued.
Wave-length.
Catalogue of Spectra.
303
Vanad um continued.
I
304
Vanadium continued.
Catalogue of Spectra.
Ytterbium continued.
305
Wave-length.
3o6
An
Introduction to
tJie
Yttrium continued.
Catalogue of Spectra.
307
ZINC.
3o8
An
Catalogue of Spectra.
Zirconium
309
continued.
Wave-length.
3IO
An
Introduction
to the
Zircon
I'Ji
APPENDIX.
On the Relative Behaviour
of the H and OF the Spectrum of Calcium.
Lines
The
We
Appendix.
and K that they became quite insignifigreatly enfeebled relatively to cant, and, if possible, disappeared altogether from the spectrum, leaving and K, or nearly so. the very simple spectrum of the two lines Professor Lockyer states that " Some of the substances which ha\e been investigated, including iron, calcium, and magnesium, have probabl)' a definite spectrum, consisting of a few lines, which can only be produced at a temperature higher than any which is at present available in laboratory experiments " {Proc. R. 5., Ixi. p. 205). In the case of calcium "(4) A spectrum consisting of the two lines at 37o6'l8 and 3737'o8 and the H and K lines, corresponding to a temperature higher than the average temperature of the spark, as before explained " {Ibid., p. 161). Such a spectrum was not actually obtained, but experiments with a large intensity coil suggested that by a still greater increase of intensity The intensity of the of the spark such a simple spectrum might appear. and K {Ibid., Table, strong blue line was reduced to one-half of
p. 162).
to the
Kayser and Runge found 106 lines of the calcium spectrum to belong system of triplets. Among the remaining lines they pointed out
;
pairs with constant differences of wave frequencies notably and K, with a difference of wave frequency of 222'9, and the more refrangible pair at 3637'o8 and 37o6'i8, with a difference of 223'i. Messrs. Humphreys and Mohler, in their experiments on the effect of pressure on the wave-lengths of metallic lines, found that in the case of and K lines were shifted only one-half as much as the calcium the blue line at 4226'9. know far too little to justify us in forming any theoretical conclusions from this peculiarity of behaviour. Indeed, there are no certain reasons why the lines of any substance should be equally shifted. It is well known that calcium, in common with nearly all substances, gives a more complex spectrum under the conditions of the arc and spark than under those of a flame. Now, in the Fraunhofer lines we have, as first shown by Kirchhoff and Bunsen, absorption spectra of the elements which correspond, speaking broadly, with those of the bright-lined spectra of the same substances as they are produced by the spark. In order, therefore, to study the modifications which the calcium undergoes in the higher regions of the chromosphere, in the prominences, and possibly in lower parts of the corona, as well as in the atmospheres of stars of different orders, it was clearly desirable that we should start with an ordinary spark spectrum. It was suggested to us strongly by the know'n rarer state of the gases in the regions above the photosphere, as well as by my long experience with the behaviour of calcium in comparison spectra, that the modifications of the calcium spectrum which we were seeking would be likely to show themselves under conditions of greatly reduced density of the calcium vapour.
We
Experiments.
For reasons which will be obvious later on, we elected to use throughout the experiments a spark of \ery small intensity. I. The break of a 6-inch Apps coil was fixed at the position of smallest acting force of the spring. So much battery power only was employed as would be just sufficient to move the break. Under these conditions, when a jar was not in connection, the feeble spark would not pass when the distance between the points exceeded i| inch.
Appendix.
2.
the experiments a jar was intercalated. 3. The same length of exposure, a very short one of a second and a half, sufficient to bring out only the strongest lines of the spectrum, was used in each experiment. in one case with 4. Two sets of similar experiments were made electrodes of platinum, and in the other with electrodes of iron. In the latter case the chief lines of iron were present with those of calcium. Methods adopted for reducing the density of the calcium vapour (a) The spark was taken between electrodes of metallic calcium. It was assumed, as was confirmed by the appearance of the spark, that with metallic calcium for electrodes the largest amount of calcium vapour would be present. (1^) The tips of the electrodes, iron or platinum, were slightly moistened with a strong solution of calcic chloride. (c) The tips were slightly washed with pure water. (d) The tips were again washed with pure water. (e) The tips were then slightly moistened with a very weak solution made by adding a drop of the strong solution to two ounces of water. Our expectations were completely confirmed. Under the conditions {a) of greatest density of the calcium vapour, when metallic calcium was employed, the blue line was as strong and possessed the same diffuse character as and K. As the density of the calcium was reduced, the lines were not found to be equally enfeebled, but, on the contrary, the blue line and the greater number of the lines were increasingly reduced in intensity relatively to H and K, until at last with the twice-washed electrodes {d) the spectrum was simplified to the condition usually existing in the prominences, in which and K only are present. now proceed to a more precise statement of the changes of relative intensity as they are presented in the photographs which accompany this paper. A. Photograph of the spark when both electrodes consist of metallic calcium. Here we ha\e present doubtless the largest amount and the The winged character of greatest density of calcium vapour. and K, of the blue line, and of the pair more refrangible than H and K, is well seen, showing that this appearance comes out when the gas is dense. If is allowed for, and the line H the greater extension of the wings of carefully distinguished from the fine lines close to it, it will be seen to possess ver)' nearly the same strength, both as regards width and length, The strength of this blue line under this as the blue line at 4226'9. condition of density is about the same as that of the line at 3737, and rather greater than the line beyond at 3706. B. Spark taken with one electrode only of metallic calcium, the other In this case the effect of a smaller density electrode being of platinum. of the calcium vapour is clearly shown in the greatly reduced wingedness of the lines. It will be remarked that the diminished density has had the these lines are now greatest influence on the pair at 3737 and 3706 much less strong than the blue line, which still holds its own, and remains The lines of the more refrangible pair are and K. about as strong as no longer diffuse at the edges. C. Spark taken between platinum electrodes moistened with a strong Here the effect of a smaller quantity of solution of calcium chloride. vapour begins to tell strongly upon the intensity of the blue line relatively and K. It may now be estimated at less than one-fourth of the to At the same time H and K have almost completely lost intensity of H. their diffuse character, and have become thinner and more defined.
In
all
We
314
Appendix.
D. The electrodes as left in the former experiment were slightly washed with pure Avater, lea\iny a trace only of calcium chloride. There is, as might be expected, an advance in the enfeeblement of the blue line and of the more refrangible pair, relatively to H and K.
SPARK SPECTRA
SHEWING EFFECT OF DENSITY ON THE RELATIVE
l>fTEVSlTIES OF TOE IIKES OP CAICIUM
K H
1S99
Fic
134-
a
\
The electrodes w ere again slightly washed with pure water, so that smaller trace of calcium chloride must have remained upon them. The enfeeblement of the blue line and of the pair has now become ery great, while and K, though thinner, remain strong. F. The electrodes were once more washed with pure water, reducin"
E,
still
Appendix.
still
further the
platinum wires.
refrangible pair thin and defined as they usually present themselves in the prominences. F*. The conditions are similar to those of F, but the reduction of density has been carried still further. H. Once more the electrodes were washed with the expectation of having removed completely the last remaining trace of calcium. To our surprise, when the photograph was developed, the lines and K came out alone. The more refrangible pair had now faded out as well as the blue line. and were now thin, and extended but a short distance in the spectrum. It must be remembered that the only condition which was varied during this set of experiments was the amount or density of the calcium vapour. The changes of relative intensity, and the modifications of the calcium spectrum produced thereby, as shown in the succession of photographs on the plate, correspond closely to the behaviour of calcium at different levels near the sim's limb, and in the atmospheres of stars of different orders. There can remain no doubt that the true interpretation of the changes in appearance of the calcium lines in the celestial bodies is to be found in the different states of density of the celestial gases from which the lines are emitted, or by which they are absorbed. similar set of experiments was made with iron electrodes. Precisely similar results as to the relative enfeeblement of the lines as with calcium chloride on platinum electrodes were obtained. Of course the iron lines were also present. As might be anticipated, in consequence of the simultaneous presence of the iron vapour, the lines of calcium were thinner than when platinum was used. Outside the range of wave-lengths which could be conveniently given on the plate, far on in the ultra-violet, there is a pair of strong lines which and K. It remains visible in photograph behave very much as when the pair at 3737 and 3706 have disappeared. This pair is situated at 3158-98 and 3179-45. It is desirable to point out again that all the photographs on the plate, and the far ultra-violet lines, were obtained with a spark of quite unusually small intensity, which was purposely made as little hot as possible, in order to emphasize the important fact that the determining condition of the spectral changes under discussion is not one of increase of temperature. In the modifications of the calcium spectrum arising from variations in the relative intensities of the lines which have been discussed in this paper, and which correspond to those observed in the celestial bodies, there does not appear to be any reason for assuming, much less any direct evidence in favour of, a true dissociation of calcium, that is, of its resolution into chemically different kinds of matter. It would be remarkable if, by decomposition through increase of temperature, a large number of lines of a spectrum should become relatively enfeebled, and that as the result of decomposition a spectrum should become simpler, and not, as analogy would suggest, more complex. It is of importance to keep in mind that the recent chemical use of the word dissociation is not equivalent to true decomposition, i.e. to a resolution of the original substance into two or more chemically different kinds of matter. It may, and does often, mean not more than a different arrangement of the parts of the molecule, while those parts are all chemias in the case of cally matter of the same kind as the original molecule the resolution of a compound molecule of peroxide of nitrogen into two identical half molecules, or in the separation of a molecule of elementary
The blue line has now practically disappeared, and the become very thin. The H and K lines have become
Appendix.
iodine into two half molecules or atoms of identical chemical characters. Such dissociations are well known, and are not of infrequent occurrence, and may, indeed, take place in connection with some of the spectral changes of a substance observed under different conditions. On the other hand, a true decomposition of a chemical element, that is, a breaking up of the molecule into simpler and quite other kinds of matter, though a notion familiar to chemists since Prout's time, and regarded as theoretically possible, is as yet unknown as a matter of fact.
Conclusions.
These experiments seem to us to furnish an adequate and consistent explanation of the behaviour of the calcium lines at and near the sun's limb. Near the photosphere, where the absorption mainly takes place by which the dark lines of the solar spectrum are formed, there would be, we should expect, a much greater density of calcium vapour than at a higher level, and we find the Fraunhofer line at 4226'9 strong, but much less broad than H and K. The recent photograph of the reversing layer and K is not reproduced there, but shows that the broad shading of probably, as Mr. Jewell concludes from his measures, lower down, where the gas is still denser, which is in agreement with photograph A on the
plate.
Higher up in the chromosphere, in the prominences, and possibly in the lower coronal regions, the decrease of the density of the gases composing them must be rapid, and the temperature gradient as determined have clearly to do, in these regions, by expansion must be rapid. with calcium vapour in a rarer state, and, except so far as the molecules may have carried up within themselves to some extent the higher heat of a lower level, or through imperfect transparency the gases may have received heat from the sun's radiation, it must be at a much lower temperature than near the photosphere. Now, the changes of the calcium spectrum which take place in these regions are those which correspond in our experiments to a very small amount of calcium vapour, and a spark
We
of small intensity. On account of the violent commotion which must exist through the strong convection currents at the sun's limb, we should not be surprised to find some calcium vapour, notwithstanding its greater density, carried high up, together with the lighter substances such as hydrogen and helium. Our experiments show how strongly the and lines may come out when only a trace of calcium vapour is present, and so, it seems to us, offer a possible explanation of the great height at which these lines may At no very great distance from the surface of be sometimes recognized. the sun the gases must become too tenuous to give a visible spectrum, and it may well be that the brilliant radiations of even very rare calcium gas at H and K may show in our instruments for some distance after the hydrogen and the other light matter associated with it have become too subtle to furnish a spectrum that we can detect. The relative beha\iour of the lines of the calcium spectrum as they present themselves in the different orders of stellar spectra, hen interpreted by the terrestrial experiments described in this paper, will throwimportant light on many of the important questions which are still pending in celestial physics. In forming conclusions as to the state of the stellar atmospheres from the different densities which may be indicated by the modifications of the calcium spectrum, it must be borne in mind that, as I have said elsewhere
Appendix.
'
conditions of the radiating photosphere and those of tlie gases above it, on which the character of the spectrum of the star depends, will be determined, not alone by temperature, but also by the force of gravity in these regions ; this force will be fixed by the starts mass and its stage of condensation, and will become greater as the star continues to condense." * It may be, though on this point we have no sufficient data, that though the stars are iDuilt up of matter essentially similar to that of the sun, the proportion of the different elements is not the same in stars which have condensed in parts of the heavens widely different from each other, or at epochs greatly separated in time. It does not seem desirable to discuss any of these questions at the present time, as we hope before long to offer some explanation of the, to some extent analogous, relative beha\iour of the lines of some other substances as observed in the sun and stars. Addendum. The following letter from our friend. Professor Liveing, which he has given us permission to publish, contains an account of early experiments on the spectrum of calcium, which not only support, by a different method of working, the conclusions of our paper, but also seem to show the possible occurrence of the line H without the line K. In our experiments, as will be seen in the spectra on the plate, both lines are always present, while the K line is stronger and longer than the line H which agrees with the photographs of the prominences taken by Hale, and
The
by Deslandres " I have been looking up some observations of Dewar's and mine on the H and K lines of calcium made in 1879. We found that when we used, for the arc, carbon poles which had been heated for two days in chlorine to remove metals, the calcium lines were not visible at first in the arc, but after a time H was seen alone and not strong after a further time K was seen, and then other calcium lines came out. No doubt the calcium had been pretty well removed from the carbon rods to some depth, but not entirely from the interior, so that as the carbon burnt away in the arc the calcium in the interior became manifest. " Again, we found that when we used a perforated pole, and passed a stream of hydrogen into the arc through it, H and K could be both entirely obliterated but by then reducing the current of gas they gradually reappeared and H always came out first, and afterwards K, and H remained stronger than K until they had both become strong and had resumed their ordinary appearance. This was seen many times. " Both observations seem to me to confirm your conclusions. In the latter case the stream of hydrogen diluted the calcium vapour, and the degree of dilution was controlled by the rate at which the gas was introduced. The mass of gas passing was too small to reduce the temperature by any considerable amount, or even, I should think, by any sensible
:
;
amount.
found also that metallic lithium introduced into the arc produced produced by hydrogen, that is, it reduced very much the strength of the H and K lines. If more than a very minute piece of lithium were introduced, the arc was invariably broken, so that we did not notice the complete obliteration of H and K with the lithium.
effects similiar to those
"We
The reduction
*
of the strength of
and
in this
case
attribute to the
Appendix.
Preliminary Note on Some Modifications of the Magnesium Line at A4481 under Different Laboratory Conditions of the Spark-Discharge.
By Sir William Hugoins and Lady Huggins.
In his "Note on the wave-length of the magnesium hne at X4481,"* Profesor Crew points out that an interesting problem still remains, namely, to discover the laboratory conditions under which the line becomes sharp,
some stellar spectra. For some years, at intervals, experiments have been made here in the laboratory on the spectrum of magnesium, with the hope of throwing light on the physical conditions of the stellar atmospheres, which we may assume to be indicated by the character of this line when present a line which in the laborator)- is subject to a very wide range of modifications, both of character and of intensity. As it may be some time before these experiments are sufficiently complete for publication, it seems desirable to reproduce at once, with this preliminary note, out of the very large number of spectra which have been taken, a few representing the most typical forms of the modifications
as in
;
of this line.
The teaching of these experiments suggests that the condition of the spark-discharge which is most potent in laringing about modifications of this line, both in intensity and in character, is the greater or less suddenTo a small extent only does the ness of the blow of the dischargee. character of the line appear to be affected by the quantity and the electromotive force of the electricity which is in action indeed, such changes as may appear are probably brought about indirectly by the larger mass of material acted upon as the discharge is made more powerful. The appearance of the line at A4481 in the spectrum at the top of the plate may be taken as representing its normal condition with capacity in When the jar is taken out of circuit, and the the secondary of the coil. discharge of the secondary takes place directly between the magnesium electrodes, the line becomes thin, defined, and of small intensity, as in spectrum No. 2. In this case the electric blows are less sudden through
;
full
researches of Schuster, Hemsalech, Schenck, Huff, and others have shown that a similar eftect follows when the jar-discharge is slowed down by the introduction into the circuit of an independent self-induction. The condition of the line in spectrum No. 3 shows the eftect of the introduction of a self-induction, the conditions of the discharge remaining otherwise the same as in photograph No. i. In spectrum No. 5 a stronger alternating current and a capacity four times as great as in No. were employed, Ijut the photograph is feeble from o\ er-exposure. On the contrary, in No. 4 the coil was excited by a feeble continuous current, and the capacity in the circuit was reduced to a small jar. The two spectra placed below were taken some years ago with another spectroscope. They are of interest in showing the great variation of intensity which the line at 4481 may undergo, without assuming its normal
i
The
Appendix.
diffused character as in No.
i.
319
spectrum No. 4 appears to be
The Hne
in
lliat in
No.
I.
320
Appendix.
have
still before us the task of the interpretation of the differences of production of the electric spark, especially as to greater or less suddenness, so as to enable us to connect them with definite conditions of temperature and density of those stellar atmospheres in which the magnesium present absorbs radiations similar in character to those photographed in the laborator)-.
We
in the
mode
London,
J(i>iuary
i,
1903.
INDEX
Aberration,
chromatic, 7 of light, 41 , spherical, 6 Abnormal solar spectrum, 166 Absorption spectra, 2, 34, 130
I
mium compounds,
,
35
delicacy of spectral reactions, 10 discovery of rubidium and coesium, 10 map of the spectra of the alkalies,
,
,
Andromeda, nebula
,
in,
88
new
star in,
107
9 Bunsen-flame spectrum,
etc.,
15, 20,
160
Angstrom and Thalen, 162 o Angstroms "spectre normal, '47 Antimony spectrum, 24
Aquila, new star in, 107 a Aquilse, 81 Arc-spectra, 39, 130, 155 Arcturus, 75, 78
Cadmium
spectrum, 24, 128, 146, 151 13, 145 Calcium spectrum, 10, 13, 21, 146, 156, 311 Carbon disulphide refractive index and
Cfesium spectrum,
Argon, 32
Argyrodite, 28 Arsenic spectrum, 32 " Atlas of Representative Spectra," 80
Stellar
of,
from
150
oxide spectrum, 164 spectrum, 28, 39, 161 Carbonic oxide poisoning, 133 Capella (a Aurigae), 75, 81, II3 Cassiopeia, new star in, 107 Castor, 81 Cauchy's dispersion formula, 55 Centre of a lens, 6 Chlorine spectrum, 33 Chlorophyll, absorption spectrum
132. 13s
of,
Chromatic aberration, 7
14,
Barium
150
bromide flame-spectrum,
Chromium
of,
salts,
absorption spectrum
chloride flame-spectrum, 14, 150 flame-spectrum, 13 iodide flame-spectrum, 15, 150 Beckley's spectograph, 5 ^ Bellatrix, spectrum of, 81 Beryllium, spectrum of, 26, 154
Berzelianite, 28 Betelgeux (a Orionis), 32, 75 Biela's comet, 1 1
35 spectrum, 28
Chromosphere, 95, 105 Chromospheric lines, 106 Clerk Maxwell, electromagnetic theory
of light, 167 Cobalt spectrum, 26, 129 Cochineal, absorption spectrum
Ccerulein, 127
of,
137
Collimator, 8
322
Comets, 113
19 Compounds, spectra of, 14, 160 Concave grating, 120
,
Index.
Fievez, map of the solar spectrum, 61 the Fizeau's method of measuring velocity of light, 41 Flame-spectra, 12 Flint-glass refractive index and dispersion, 7 Fluoresceine, absorption spectrum of,
spectra
of,
17
oxide spectrum, 17 spectrum, 17, 129, 148 Corallin absorption spectrum, 137 Cornu, determination of velocity
light,
,
137
Focus of a
of
lens, 5
42
of solar spectrum, 61
the
map
Fraunhofer
lines, 49,
59
Borealis,
new
8
dis-
Crown
Cyanogen band
64
spectrum, 164 Cygni /3 a double star, 76, 113 Cygnus, new star in, 106
Density, influence
of,
on the spectrum,
'55,3"
Deslandres' formula for band-spectra, 161 Dibbits' carbon spectrum, 161 Didymium absorption spectrum, 35
Diffraction grating, 45, 120, 174 spectrum, 41, 47 Discontinuous spectra of gases, 1 Dispersion of light, 3 DiNading engine for ruling gratings, 120
H.E;MOGLOiilN, absorption spectrum of, 132 Hale's spectroheliograph, 165 Halley's comet, 113 Hartley method of photographing
spectra, 125
spectnmi
137
diffraction grating, 174
79
Echelon
line
,
4481
.31S
lines, 91,
, ,
on the nature of the nebulae, 82 on the spectra of comets, 117 Hydrogen spectrum, 31, 89
,
series in,
138
of,
compound spectrum
139
Erbium salts, absorption spectra Erythrosine, 127 Evolution of the stars, 75, 76
FABK.Y and
of,
36
IMPURITV
39
Indium spectrum,
Induction
of
coil,
24, 154
iS
Index.
Interpolation curve, 49-51 formulEE, Cauchy, 55
323
Hartmann,
field,
,
171
lines,
wave-lengths of cadmium
lines,
'57 Micrometer, binlar, 53 bright-line, 53 , divided-lens, 53 , tangent-screw, 52 Microspectroscope, 134 Mills spectrograph of the Lick Ob,
series
formulrc,
Minimum
Mira
servatory, 93 deviation,
(0 Ceti),
map
Krypton discovery, 32
spectrum, 33
Nebul/e, 82
,
Nebular hypothesis, 84 Neon, 32 Nickel spectrum, 26, 131 Nitrogen peroxide absorption spectrum, 36 Nitrogen spectrum, 32
Novre, or
new
stars,
107
centre of
a,
6
7 18,
,
,
of,
82
20
Light, electro-magnetic theory , interference of, 44 , velocity of, 41, 42 Lithium flame-spectrum, 12
of,
167
68
Oscillation frequencies, 48, 57 Oxygen, absorption lines of
spheric,
,
atmo-
68
of,
142 , series in the spectrum of, Liveing and Dewar, carbon spectrum, 162 Lockyer, carbon spectrum, 162 dissociation of the elements, 91 Long and short lines, 20, 158 Lorentz, explanation of the "Zeeman
,
spectra
32
Palladium
of,
35'
13
Perseus,
new
star in,
108
effect,"
of the group
Mapping a spectrum, 9
Magenta, absorption spectrum
of,
137
17,
22
Mercury
light,
Cooper-Hewitt, 24
161 Pole-star a spectroscopic binary, 113 Pollux, 75 Potassium flame-spectrum, 13 series in the spectrum of, 144 Potsdam spectrograph, 105 Pressure, influence of, on the spectrum,
,
155
3 24
Prism,
,
,
Index.
I
Procyon, 75, Si Prominences, 96, lOI Protyle, 89 ^ Puppis, 89 Purpurin absorption spectrum, 136
Quartz
lenses and prisms, tlieir use in photographing the spectrum, 71, 123
Spark- and arc-spectra, 39, 128, 129, 131. 158 Spark, phenomena of the, 159 Spectra, relationships between, 141 Spectrograph, 105 Spectroheliogiaph, 165 Spectrometer, 122 Spectroscope, 8, 12, 59 Spectroscopic binaries, 112
Spherical aberration, 6 Standard wave-lengths, 185 Stars and Sun, temperature of the, 92
Stellar spectra, 72, 75, 78
in spectra, 151
Red
stars,
75
Regulus, 75 Relationships in spectra, 138 Reflection, law of, 43 Refraction, law of, 3, 43 Refractive index, 3 Resolving power of a spectroscope,
121, 174, 178 Rigel, 75, 81
Strontium spectrum, 14, 146 Sulphur spectra, 28 Sun, elements present in, 59 chromosphere, its photosphere, corona, and prominences, 95 Sun-spot spectrum, 102, 103
,
Sun-spots, loi, 102 Swan spectrum, 20, 160, 161 Swift's comet, 116
measurements, 157 concave grating, 45, 120 map of the solar spectrum, 61
, ,
Telluric
185
of
magnetic
calculation
field,
lines of the solar spectrum, 67, 68 Tellurium spectrum, 28 Thallium spectrum, 2, 24, 128, 148 Thiele's formula for band-spectra, 161 Thomson on "corpuscles," 90 Tin spectrum, 24, 12S
and
Precht,
of
spectra, 154
rule,
153
of,
137
Types of
stellar spectra, 75
Safranin'e, absorption
Sagittarius,
spectrum
106
of,
new
star in,
Vacuum
tubes, 31
and Rydberg's rule, 153 Secchi types of stellar spectra, 75 Selenium, spectrum of, 28
Self-induction, 159 Sensitiveness of photographic plates for different parts of the spectrum, 74
Silicates, treatment of, 13,
26
Water,
sion, 7
refractive
motion
of,
94
Slit of spectroscope, S
\Vater-vapour in the atmosphere, 68 Wave-length, determination of, 47 White stars, spectrum of, 76
Index.
Wine, adulteration of, 137 Wolf-Rayet stars, 78
Wollaston, carbon spectrum, 161
325
stars,
Yellow
74
Xenon,
52
" Zeeman effect," 168 Zinc spectrum, 24, 128, 145 Zirconium spectrum, 28
THE END.
PKINTED
r.V
INDEX OF SPECTRA
With a Preface by
Introduction
Sir H. E. Roscoe, F.R.S., D.C.L., etc., and an on the Methods of Measuring and Mapping Spectra. Revised Edition, greatly Enlarged. Abel Heywood & Son, Manchester. May be obtained from the Author at
the
following
;
prices, post
free:
First
Edition
(1872),
with
Enlarged Edition (1889), bound volume, including Appendix A, ^i is. The two Editions, bound together,;^i 7^-.
plates, 6s.
And
Ajipendix B (Table of corrections to reduce to Rowland's standard, ultra-violet Spectra of Co and Ni, etc.), 3^-.
Appendix C (Spectrum of
Hydrogen),
5^.
Appendix D (Arc-spectra of Ga, Li, Na, K, Rb, Cs, Mg, Ca, Zn, Hg, absorption spectrum of Bromine, arc-spectrum of Alumina), Appendix E (Spectra of Air, Cu, Ag, Au, Al, In, Tl, C, CN, N, and the Table of reductions to vacuum), 3^. 6d. Appendix
Appendix
Cd,
Cd, Ba,
Si,
ammonia,
Li,
(the arc-spectra of Cr, Sn, Pb, Sb, Bi, flame-spectra of etc.), 2s. 6d.
K, Na,
etc.,
(Rowland's Standard Wave-lengths, spectra of Mercury, Helium, and oxy-hydrogen spectra), 4s.
the
Appendix PI
(the three spectra of Argon, the arc-spectrum of Titanium, spark-spectra of Copper, Silver, and Gold), 4s.
I
Appendix
(the arc-
4s.
Appendix J
and spark4^.
spectra of Platinum),
Appendix
Appendix
K
L
of Uranium),
Appendix M (the arc-spectrum of Manganese and Vanadium, the spark-spectrum of Gold and Silicon, and the spectrum of Argon), 5^-. Appendix N (Flame-spectrum of Radium, infra-red spectra of the Alkalies, ultraviolet
spectrum of Thorium),
3j'.
Appendix
(Arc-spectrum of Molybdenum, spark-spectra of Calcium, Scandium, Indium, Beryllium, Lithium, ThaUium, Potassium, Rubidium, Cfesium,
3^.
NOTE. Appendices B
bound
in
one volume,
"SHIRLEY,"
166,
A SELECT
LIST OF
IN
BOOKS
Messrs.
CO.
E.G.
PAGE
....
.
-
38
9 35
14 20 33 32
....
26
8
ig
19
Bacteriology Biology
Science
23
18
Botany
BuiLDiNO Construction
-
34
-
Naval Architecture
-
Calculus Chemistry
Conic Sections
-
Navigation 10 Optics
14
2
.20
12
-
Photography
Physics
12
5
10
. Physiography .22 16 Physiology Electricity 32 Elsmfntary Science Manuals- 3S Practical Elementary Science Series Engineering 17 40 11 Proctor's (R. A.) IVorks - 21 Euclid Gardening 35 Sound 13
Dynamics
6
-
15
17
8
Heat
-
Hydrostatics 6 Light 13 10 Telephone Logarithms London Science Class- Books - 40 Text-Books of Science Thermodynamics Longmans' CrviL Engineering i8 Trigonometry Series Machine Drawing and Design 19 Tyndall's (John) Works 16 Veterinary Medicine, etc. Magnetism Manufactures 24 Workshop Appliances 6 Zoology Mechanics
-
24 17
-
17
37
13 13
36
31
iS
32
Scientific
Co.
.J7J'Z)]'/]/^i^':
A
Si. net.
Manual
,
ot
Ouantitntive Analysis for Students of Agriculture. With 49 Illustrations. Crown 8vo., B.Sc. Lond., F.I.C.
By FRANK
T.
AddYMAN,
German
9.1.
Edition by
With
58
8vo,
6</.
net.
COLEMAN
Frank
8^'0.
,
AND ADDYMAN. Y^KdlCKl. AGRICULTURAL CHEMISTRY. By Bernakd Coleman, A.R.C.Sc., F.I.C, and
J.
T.
Addvman,
With 24
Illustrations.
Crown
ij.
kd. net.
William Crookes, F.R.S., etc. Third With 67 Woodcuts. 8vo., 2ii. net. FURNEAUX.YX.'^l.lYM'YK^Y CHEMISTRY, Inorganic and Organic. By W. FuRNEAUX, F.R.G.S. Lecturer on Chemistry, London School Board. With 65 Illustrations and 155 Experiments. Crown 8vo. 2J. td.
chiefly Inorganic,
Ci?(9CA'.5. ANALYSIS,
Edition, Rem-itten
and Enlarged.
F. C.
Garrett, M.Sc.
Durham
Dunelm.), Assistant Lecturer and Demonstrator in Chemistry, the and Arthur Harden, College of Science, Newcastle-on-Tyne M.Sc. (Viet.), Ph.D., Assistant Lecturer and Demonstrator in Chemistry, Manchester University. With 19 Illustrations. Crown 8vo. 35.
;
,
JAGO.WoxVi
by W.
JAGO,
F.C.S., F.I.C.
INORGANIC
CHEMISTRY,
an Introduction
THEORETICAL
to the Principles of
AND
PRACTICAL. With
AN INTRODUCTION TO PRACTICAL INORGANIC CHEMISTRY. Crown 8vo. u. INORGANIC CHEMISTRY, THEORETICAL AND
,
e,/.
PRACTICAL.
8vo.
,
A Manual
for
With Plate
6d.
KLOCKER.
tory
a LaboraHandbook, By Alb. KlScker. Translated by G. E. Allan, B.Sc, and J. H, Millar, F.I.C. With 146 Illustrations in the te.xt. 8vo., I2j. net.
:
FERMENTATION
J.
ORGANISMS
With
Work.
Scholar,
chester.
W. Mellor, D.Sc,
;
New
Zealand University Research Fellow, the Owens College, ManWith 142 Diagrams. 8vo. I2j. 6c;'. net.
,
by D.
MENDELEliFF,
St. Petersbura;.
Professor of
University of
Russian (Sixth Edition) by George Kamensky, A. R.S. M., of the Imperial Mint, St. Petersburg; and Edited by T. A. Lawson, B.Sc., Ph.D. Fellow of the Institute of Chemistry. With 96 Diagrams and Illustrations 2 vols.
Kamensky '
'
Scientific
is"
Co.
il/^
F^i?. OUTLINES OF
Williams,
B.Sc.
8vo., gf.
THEORETICAL CHEMISTRY.
By LoTHAR Meyer, Professor of Chemistry in the University of Tubingen. Translated by Professors P. Phillips Bedson, D.Sc, and W. Carleton
J/ZZZ^i?. INTRODUCTION TO THE STUDY OF INORGANIC CHEIVIISTRY. By W. Allen Miller, M.D., LL.D. With
71 Illustrations.
Fcp. 8vo.
3J. (3d.
MUm.A COURSE
M.
Part
P.
,
OF PRACTICAL CHEMISTRY. By
Prselector in Chemistry of Gonville
(3 Parts.)
M.
and Caius
College, Cambridge.
I.
Elementary.
Intermediate.
Part II.
Part III.
JVEWTIf.Works by G.
strator in the
NEWTH,
F.I.C.,
F.C.S.,
Demon230
CHEMICAL
Illustrations.
LECTURE
Crown With 100
8vo., 6j.
EXPERIMENTS.
Crown
8vo., 6s. 6d.
With
Illustrations.
155
With
io8
Crown
8vo.
is.
6d.
Mollwo
Chemistry Department, Borough Polytechnic Institute, 8vo., 3s. 6d. Illustrations and Spectrum Plate,
FZIMMJ?.
Plimmer.
THE
Aders
CHEMISTRY
FOR
Introductory.
Fcp. 8vo.
,
Fcp. 8vo.,
2S. 6d.
for Acids.
Illustrations.
Crown
8vo., 2s.
6d.
PRACTICAL
INTRODUCTION TO CHEMISTRY.
,
Intended to give a Practical acquaintance with the Elementary Facts and With 25 Illustrations. Crown 8vo. 2j. Principles of Chemistry.
Scientifu
6^"
Co.
SMITH
CH E M ISTRY Conlinued. AND HALL.Tn'E. TEACHING OF AND PHYSICS IN THE SECOiNDARY SCHOOL.
CHEMISTRY
By Alexander
TEXT-BOOKS OF PHYSICAL CHEMISTRY. Edited by Sir William Ramsay, K.C.B.,F.R.S THE PHASE RULE AND ITS APPLICATIONS. By
Smith, B.Sc, Ph.D., Associate Professor of Chemistry in the University of Chicago, and Edwin H. Hall, Ph.D., Professor of Physics in Harvard University. With 21 Woodcuts, Bibliographies, and Index. Crown 8vo., 6s. net
FiNDLAY, M.A., Ph.D., D.Sc, Lecturer and Demonstrator in Chemistry Uni^erslty of Birmingham. With ii8 Figures in the Te-\t, and an Introduction TO Physical Chemistry by Sir Wm. Ramsay, K.C.B.,F.R.S Cr 8vo i;j
Alex.
By
I.
Sir
ELECTRO-CHEMISTRY.
Crown Part II.
Part
Lehfeldt, p.Sc, Professor of Physics at the East London Technical CoUege Including a Chapter on the Relation of Chemical Constitution to Conductivity ^'^''" ^'^^'^'" '" "^<= University of Birmingharn'.
General Theory.
By R.
Svo?""'
J
7-,
,
Applications
nepoUouHiig
to
I
Electrolysis,
Batteries, etc.
By
. ^In preparation.
..
.-
SPECTROSCOPY.
F.I.C.
!;.
C. C. Baly,
By F. G. DonRELATIONS BETWEEN CHEMICAL nan, M.A., Ph.D. CONSTITUriON AND PHYSI- CHEMICAL DYNAMICS AND RECAL PROPERTIES. By Samuel ACTIONS. By W.
Smiles, D.Sc.
J.
THERMODYNAMICS.
Mellor,
D.Sc.
THORNTON and
ANALYSIS. By Arthur Thornton, M.A., and Marchant Pearson, "= Bradford Grammar School. Medium 8vo., 2. B.A., TZT^^T^ ,',?"=, ^'f '''' by T. E. THORPE, C.B., D.Sc. (Vict.), Ph.D.,
PEARSON.V.OTYJS,
ON VOLUMETRIC
Laboratory,
^ /f
cuts.
A DICTIONARY OF APPLIED CHEMISTRY Vols. and each. Vol. III., 63J. ^TT.^!;'^QUANTITATIVE CHEMICAL ANALYSIS. With
I.
London
-i
vols
THORPE
Fcp. 8vo.,
88 Wood-
4J. 6<;.
AND J/^/i?. QUALITATIVE ALYSIS and LABORATORY PRACTICE. ByT. CHEMICAL ANE Fhorpe, C B. Ph a. """' ^'-^^ ^'"^ "^'^ "^ ^^^'^ ^^ 57 IUusfr^i^ns.^-^c^.-8'^o Ta^"^
A. TILDEN, D.Sc. London, of Chemistry in the Royal College of Science,
WILLIAM
Cro^Svo
Ztls^^'^''^'Analysis.
^'^-
'-
and Systemadc
''
^'"^
^""^^"^^^^oTZe^rTp:
principles
ChpVr7
.
PRACTICAL CHEMISTRY.
Fcp. 8vo.,
li. 6./.
The
of '
On.Hf Qualitative
Scientific
Co.
CHEWIISTRY eoWKerf.
Rewritten by H. Forster Morley, M.A., D.Sc. Fellow of, and lately Assistant Professor of Chemistry in, University College, London and M. M. PATTISON Muir, M.A., F.R.S.E., Fellow, and Praelector in Chemistry, of Gonville and Caius College, Cambridge. Assisted by Eminent Contributors. 4 vols. 8vo., i net.
WHITELEY.^oxY^
by
R.
LLOYD WHITELEY,
With
F.I.C.,
West Bromwich.
Explanatory Notes,
CHEMICAL CALCULATIONS.
Crown
8vo., 2j.
Problems and Answers, specially adapted for use in Colleges and Science Schools. With a Preface by Professor F. Clowes, D.Sc. (Lond.), F.I.C.
ORGANIC CHEMISTRY;
Illustrations.
the Fatty
Compounds.
With 45
Crown
8vo.
y.
6d.
PHYSICS, ETC.
Headmaster
IS.
of the
Book
Book
I.
Elementary
6d.
By John Bidgood, (In Three Books.) Gateshead Secondary Day School. With 120 Illustrations. Crown Physics.
Bvo.,
II.
Physics
Bvo.,
is.
and Chemistry.
With
122
Illustrations.
Crown
6d.
Book III. Chemistry. With 108 Illustrations. Crown 8vo., 2s. ^C.S.^. RESPONSE IN THE LIVING AND NON-LIVING.
By JAGADIS Chunder Bose, M.A. (Cantab.), D.Sc. (Lond.), Professor, dency College, Calcutta. With 117 Illustrations. 8vo., los. 6d.
Presi-
GANOTWorks
.
ELEMENTARY
Translated and by PROFESSOR GANOT. Edited by E. Atkinson, Ph.D., F.C.S., and A. W. Reinold M.A., F.R.S. TREATISE ON PHYSICS, Experimental
and Applied. With 9 Coloured Plates and Maps, and 1048 Woodcuts, and Appendix of Problems and Examples with Answers. Crown 8vo. 15^.
,
With
7 Plates, 632
Woodcuts, and an
Crown
GLAZEBROOK
R. T.
Illustrations.
and
By
134
By
Guthrie, Ph.D.
With
91 Diagrams.
Fcp. 8vo.,
is.
6d.
SCIENTIFIC
With
Translated by E. ATKINSON,
Ph.D., F.C.S., formerly Professor of Experimental Science, Staff College. 68 Illustrations. 2 vols. crown Bvo. y. 6d. each.
By
John
Road
Crown
2s. bd.
Scientific
Co.
PHYSICS, ^tCConlinued.
yl/^CZ^^iV. EXERCISES IN
NATURAL PHILOSOPHY.
GASES.
ElemenEmil
By Dr. OsKAR
By Magnus Maclean, D.Sc, Professor of Electrical Engineering at the Glasgow and West of Scotland Technical College. Crown 8vo., 41. bd.
P/EYjEJ?.TRE KINETIC
tary
,
THEORY OF
Appendices.
FAJV
IN
a Preface by JOHANNES WiSLiCENUS, Professor of Chemistry at the University of Leipzig and an Appendix Stereochemistry among Inorganic Substances,' by Alfred Werner, Professor of Chemistry at the University of Zurich. Crown 8vo., 6s. 6d.
Laboratory
Illustrations
With 120
and
A TEXT-BOOK OF PHYSICS.
Illustrations,
and a Collection
8vo.,
los.
of
Large crown
6d.
By A, M.
4s. 6i.
C.B., F.R.S.
With
Illustrations.
Crown
Svo.
W/i?/(?^r. ELEMENTARY
,
PHYSICS.
By
Mark
R.
Wright, M.A. Professor of Normal Education, Durham College With 242 Illustrations. Cro^vn Svo. 2J. 6d,
of Science.
S.
Fcp. 8vo.,
ts. 6d.
GOOL)F. Works
GOODEVE,
Normal School
M.A.,
formerly
of Science,
and
With 357
lUustra-
PRINCIPLES OF MECHANICS.
With 253
:
Illustrations
and
A MANUAL OF MECHANICS
for
numerous Examples.
Crown
Svo.
6j-.
an
Elementary Text-Book
Students of Applied Mechanics. With t38 Illustrations and Diagrams, and 18S Examples taken from the Science Department Examination Papers,
with Answers. Fcp. 8vo.,
2j. 61/.
GCOZ>yl/^iY. MECHANICS
Crown
Svo.
,
By John Goodman, Wh.Sch., M.I.C.E., M.I.M.E., Professor of Engineering in the University of Leeds. With 714 Illustrations and numerous Examples
gi. net.
APPLIED TO ENGINEERING.
Scientific
is'
Co.
GJiI F. LESSONS
By W. H. Grieve,
School Board,
Stage
I.
IS.
IN
ELEMENTARY MECHANICS.
R.N., Science Demonstrator for the London
late Engineer,
etc.
With
6d.
number
bd.
of Examples.
Fcp. 8vo.,
Stage
Stage
2. 3.
is.
is.
6d.
MAGNUS.--\Noi\s
B.Sc, B.A.
Introductory
to the study of Physical Science. Designed for the Use of Schools, and of Candidates for the London Matriculation and other Examinations. With numerous Exercises, Examples, Examination Questions, and Solutions, etc., from 1870-1895. With Answers, and 131 Woodcuts. Fcp. 8vo. 3J. 6d.
,
Key
or,
5J. -^^d.
Fcp. 8vo.,
2s.
is. 6d.
The Worked
:
F17LJV. MECHANICS
Theoretical,
mental. By W. W. F. PULLEN, Wh.Sch., M.I.M.E., A.M.I.C.E. Crown Bvo., 4^. 6d. 318 Diagrams and numerous Examples.
JiOBIJVSOJV. ELEMENTS
Statics).
OF DYNAMICS
for
and
By
the
With numerous Exercises. A Text-book Rev. J. L. Robinson, M.A. Crown 8vo., 6s.
Junior Students.
SM/TJI. Works
by J. HAMBLIN SMITH, M.A. ELEMENTARY STATICS. Crown 8vo., 3^. ELEMENTARY HYDROSTATICS. Crown 8vo., ^s. KEY TO STATICS AND HYDROSTATICS. Crown 8vo., 6s. TARLETON.AN INTRODUCTION TO THE MATHE-
College,
Crown
TAYLOR. WorVs by J. E. TAYLOR, M.A., B.Sc. (Lond.). THEORETICAL MECHANICS, including Hydrostatics
Pneumatics. With 175 Diagrams and Illustrations, Questions and Answers. Crown 8vo. zs. 6d.
,
and
Papers,
etc.
With 122
Illus-
amination Papers,
Crown
8vo.
2j.
6d.
By Arthur Thornton, M.A., Including Kinematics, Statics and Kinetics. With 200 Illustrations, 130 Worked Examples, and over goo F.R.A.S. Examples from Examination Papers, etc. Crown Svo. 4^. 6d,
Scientific
(Sf
Co.
TWISDEN.^ oxY%
their Study.
by the Rev.
;
JOHN
F.
TWISDEN, M.A.
Crown
PRACTICAL MECHANICS
With 855
8vo.
,
an Elementary Introduction to
and 184 Figures and Diagrams.
Exercises,
lor. 6rf.
THEORETICAL
MECHANICS.
With
Crown
8vo.
,
172
is. 6rf.
Examples,
fr/ZZ/.^^5C> TV. INTRODUCTION TO THE MATHEMATICAL THEORY OF THE STRESS AND STRAIN OF ELASTIC
SOLIDS.
By Benjamin Williamson, D.Sc, F.R.S. Crown
8vo.,
sj.
WILLIAMSON
AND
TAELETON.Mi ELEMENTARY
Containing Applications to Thermodynamics,
TREATISE ON DYNAMICS.
with numerous Examples. By Francis A. Tarleton, LL.D.
W0RTIIINGT0N.~V>Y-^NM1CS> OF ROTATION:
mentary Introduction to Rigid Dynamics.
F.R.S.
Crown
By
F.
G.
CAT'F^i?^-. ELEMENTARY
,
MENSURATION.
Chivers, Head Master of H.M. Dockyard School, Pembroke. to the Examples. Crown 8vo 51.
By G. T. With Answers
By the Rev. Alfred Hiley, MA. With a Chapter on Land Surveying by the Rev. John Hunter, M.A. Containing numerous Examples, and embodying many of the Questions set in the Local Examination Papers. With Answers. i2mo. 2J. 6t/.
,
Key,
it.
Z(9i9G^. MENSURATION
Alfred Lodge, M.A.,
Professor of Cooper's Hill.
late
By
Fereday Fellow of St. John's College, Oxford; Pure Mathematics at the Royal Indian Engineering College,
With Answers.
By
Alfred
J.
ScienHfic
Careen,
<^ Co.
To meet
the Requirements of the Oxford and Cambridge Junior Local Examinations, the College of Preceptors, etc. By W. S. Beakd, F.R.G.S., Head iVIaster of Fareham Modern School. With Answers to Exercises and Examination Papers. Crown 8vo. ij.
By Arnold Lupton, Mining Engineer, Certificated Colliery Manager, Member of the Institution of Civil Engineers, etc. With 216 IllusMedium 8vo., 12J. net.
;\^^5'^/r. PRACTICAL
Illustrated
MENSURATION.
RULE.
By
By A. Nesbit.
izmo.,
3J.
td.
Key,
5j.
G.
L.
Smith.
IJ.
net.
ALGEBRA, ETC. ^ Co.'s Catalogue of Educational and School Books, ANNALS OF MATHEMATICS. {Published under the
see
Longmans
Issued Quarterly.
CONSTABLE
2J.
ALGEBRA UP
2s.
Mills, B.A.
Crown
each.
8vo.
By W. G. CONWith Answers,
Three
Crown
Answers.
Crown
Svo.,
paper covers,
By A. G.
to
With Answers
the
By William
Crown
S.
Modern
School, Fareham.
8vo.,
7l/.ZZ0^. HIGHER MATHEMATICS FOR STUDENTS OF CHEMISTRY AND PHYSICS. With special reference to Practical
Work.
By
J.
Manchester.
Fellow,
6d. net.
The Owens
College,
SMITJI. Works
num
,
3J.
6d.
on Sale, price
EXERCISES IN ALGEBRA.
2s.
With Answers.
Fcap.
8vo.,
6d.
Copies
may
By
,
formerly Fellow of Gonville and Caius College, Cambridge, and C. H. P. MAYO, M.A. formerly Scholar of St. Peter's College, Cambridge Assistant Masters at Harrow School. Crown Svo. 35. td. or
W. Welsford, M.A.
;
with Answers,
4J. 6d.
10
Scientific
containing
Geometric
<ind
^7l//ri7: GEOMETRICAL
Hamblin Smith, M.A.
Crown
CONIC SECTIONS.
Svo.,
By
J.
y. bd.
GKAVniCA-L
,
CALCULUS.
By
Arthur
Barker, B.A., B.Sc. With an Introduction by John Goodman, A.M.I.C.E. With 61 Diagrams. Crown 8vo. us. 6d.
H.
MURRAY.~\NoxYs
Ph.D.
by
AN INTRODUCTORY COURSE
EQUATIONS.
Crown
8vo.
8vo., +1.
6d.
DIFFERENTIAL
INFINITESIMAL CAL-
y.t.
6</.
O'DEA.-A-H
By James
J.
ITHMS, EXPONENTIAL AND LOGARITHMIC SERIES, UNDETERMINED CO-EFFICIENTS, AND THE THEORY OF DETERMINANTS
O'Dea, M.A.
INTEGRAL CALCULUS.
OF THE
By Thomas Tate.
DIFFERENTIAL AND
i2mo.,
^s. bd.
TAYLOR. Works by F. GLANVILLE TAYLOR. AN INTRODUCTION TO THE DIFFERENTIAL AND INTEGRAL CALCULUS
ovo.,
Cr
gs.
Answers
to Exercises.
Crown
MENSUR.ATION
With
iv.
6d.
fVLLLLAMSOiV.W'oxks by
^^N
BENJAMIN WILLIAMSON,
^
D.Sc.
"""^^^"^
"^^
a4
CA^,^m!^^^^^^'
Crown
Svo"; ro.
6./
a ChaDlv'on'r.'"r^
^^ '"'
' ^'^-''^''^^^
Scientific
Works published
by Longmans,
Green, &'
Co.
ii
GEOMETRY AND
** Forother
Works,
sec
EUCLID.
and School Books.
Longmans if
FROM THALES TO
bd.
Allman.
8vo.
los
By Frank R.
Part Part
I.,
II.,
being tlie subject-matter of Euclid, Book I, Crown 8vo., is. containing the subject-matter of Euclid, Book III. 1-34, and
Book IV.
4-9.
Crown
8vo.
i.f.
Section I. complete. Crown 8vo. 2.f. Section II., containing the remainder of Euclid, Books III. and IV,, together with the subject-matter of Books II. and VI. With explanation of Ratio and Proportion, Trigonometric Ratios and iVIeasurement of Circles. Crown 8vo. xs. 6d. Sections I. and II. in one volume. Crown 8vo., 3^. 6d. Section III., containing the subject-matter of Euclid, Book XI., together with a full treatment of volume and surface of the cylinder, cone, sphere, etc. Crown
,
8vo.,
I.f.
6d.
Sections
I., II.
and
III.
complete
in
one volume.
Crown
8vo.,
4.?.
6d.
CASEY. \Nox\is by JOHN CASEY, LL.D., F.R.S. THE ELEMENTS OF EUCLID, BOOKS L-VL
positions, I.-XXT. of
and Pro-
Cone,
45. 6d.
etc.
Key
I.
Fcp. 8vo.
6s.
Part
I.
Fcp.
8vo., 3J.
6d.
AND CONIC
SECTIONS.
8vo.
,
Containing
an Account of
Crown
12s.
By
the
M.R.I. A.
Edited by
Charles
Tenths and
Five-fifths.
Centimetres.
2.
LOW'S IMPROVED SET SQUARES. A I 45 4" 1 fB I 45 4" each i/- net. A i A2 1/3 A2 4S6" -or{B2 4S6" A3 A3 4S6i"J Us 45 84" W' 3. LOW'S IMPROVED PROTRACTORS
,, ,,
Designs A & B. 45 to 60. 60 4" 1 TB i 60 4" each i/- net. 60 6" Ur-^ B 2 60 6" ,, 1/3 ,, 60 8}/' ) [B 3 60 8I" 2/-
(Celluloid).
3.
Protractor No.
2.
3"
in
radius,
marked
Protractor No.
^degrees, gd.
4.
net.
2.t.
5.
6d. net. LOW'S ADJUSTABLE PROTRACTOR SET SQUARE. LONGMANS' BLACKBOARD ENGLISH AND METRIC RULE. One
Metre; marked
inches.
2S. 6d.
and quarter-
Scientific
AND
DE-
By David Allan Low, Professor of EngineerCrown 8vo. ing, East London Technical College. With 64 Figures, y. Part II. With 114 Figures, 2s. Part I.
the Areas of Irregular Figures and specially useful for determining the Mean Effective Pressure from Indicator Diagrams from Steam, Gas and other Engines. Designed by D. A. Low. With Full Instructions for Use. is. net.
SOLID GEOMETRY.
Hammond Morris
Crown
8vo., 2j. 6d.
^/C^^/5. GEOMETRICAL DRAWING FOR ART STUDENTS. Embracing Plane Geometry and its Applications, the Use of Scales, and the Plans and Elevations of Solids as required in Section I. of Science Subjects. By I. Hammond Morris. Crown 8vo. zs.
,
J. to 6, and portions of Books ii and 12, of Cr. 8vo., 3J. 6d. Key, crown 8vo., %s. 6d.
By
Hamblin
Books
and
limp cloth,
i^. 6^,,
may be had
separately.
SFO ONER, THY. ELEMENTS OF GEOMETRICAL DRAWElenaentary Text-book on Practical Plane Geometry, including an Introduction to Solid Geometry. Written to include the requirements of the Syllabus of the Board of Education in Geometrical Drawing and for the use of Students preparing for the Military Entrance Examinations. By Henry J. Spooner, C.E., M.Inst.M.E. Crown 8vo., y. 6d.
ING: an
^/Z:567V: GEOMETRICAL
H^/iVr.fi'j?. By S. H. Winter.
tion of Scales, the
DRAWING.
For
the
,
use
of
Candidates for Army Examinations, and .is an Introduction to Mechanical Drawing. By W. N. Wilson, M.A. Parts I. and II. Crown 8vo. 45. 6(i. each
Use
Including Practical Plane Geometry, the ConstrucI. of the Sector, the Marquois Scales, and the Protractor.
Post 8vo.,
55.
With
3 Plates
ON PHOTOGRAPHY.
OPTICS.
By Sir William
DRUDE.THE THEORY OF
By Paul Drude,
Professor of Physics at the University of Giessen. Translated from the German by C. RiBOEG Mann and Robert A. Millikan, Assistant Professors of Physics at the University of Chicago. With no Diagrams. 8vo., iS-f. net.
GLAZEBROOK.VnYSlCK'L OPTICS.
Woodcuts
of Apparatus, etc.
By R. T. GlazeWith 183
a Practical
6s.
:
VANDERPOEL.CO'LO'R. PROBLEMS
for the Lay Student of Color. Plates in Color. Square Bvo.
,
Manual
With 117
net.
:
a Treatise
on the Use
of the Lantern in Exhibition and Scientific Demonstration. By Lewis Wright, Author of Light a Course of Experimental Optics '. With 232 lUustratipns
Crown
8vo., 6;.
Scientific
Co.
13
F, R. S.
late
U niversity
of Ireland.
By
A. Clement Jones, M.A., Ph.D., late Open Scholar and Senior Hulme Exhibitioner of Brasenose College, Oxford Senior Mathematical Master of Bradford Grammar School. Crown 8vo., 2S.
Crown
8vo.
4^. 6rf.
Key,
7s. 6d.
SOUND, LIGHT, HEAT, AND THERMODYNAMICS. DEXTEJi. 'ELEMENTARY PRACTICAL SOUND, LIGHT
AND HEAT.
Technical
Illustrations.
By Joseph S. Dexter, B.Sc. (Lond.), Physics Master, Day School, The Polytechnic Institute, Regent Street. With 15^ Crown 8vo., 2S. 6d.
EMTAGE.EIGWT.
By W. T. A. Emtage, M.A.,
With 232
Illustrations.
Director of
8vo., 65.
Crown
AS A
By Her-
By
J.
Clerk Maxweli,
Ray
M.A., F.R.SS. L. and E. With Corrections and Additions by Lord LEIGH. With 38 Illustrations. Fcp. 8vo. 4J. 6d.
,
By Dr.
Max Planck, Professor of Theoretical Physics in the University of Berlin. Translated, with the Author's sanction, by Alexander Ogg, M.A., B.Sc, Ph.D. ,late 1851 Exhibition Scholar, Aberdeen University; Assistant Master,
js, 6d. net.
By
J.
Hamblin Smith,
8vo.
,
Crown
3J-.
by
JOHN TYNDALL,
Fcp. 8vo.,
\s.
D.C.L,
F.R.S.
WRIGHT.\Yox\i%
Illustrations.
MARK
2S.
and
6d.
ADVANCED HEAT.
With
numerous
14
Scientific
Co.
Rivingtons'
8vo.
,
'.
With 385
Illustrations.
Crown
4J.
^^iVA'; STYLE
With 102
Plates
IN
by
FURNITURE.
C.
By R. Davis
net.
Benn.
W.
Baldock.
8vo., 21J.
^C(9A''7?. ELEMENTARY PRACTICAL BUILDING CONSTRUCTION. By F. W. Booker. With Illustrations. Crown 8vo., 2j. arf.
By Edward
J.
London.
Crown
8vo.
is.
6d.
Wyatt Papworth.
Bv A. D.
F.
Revised (1888), with Alterations and Considerable With 1700 Engravings. 8vo. zxs. net.
,
Hamlin, A.M.
With 229
Illustrations.
Crown
PARKER
AND
:
HOME A
Raymond Unwin.
Collection of Lectures and Illustrations. By Barry Parker With 68 Full-page Plates. 8vo. io.t. 6d. net.
By
H. W. Richards, Examiner in Brickwork and Masonry to the City and Guilds of London Institute, Head of Building Trades Department, Northern Polytechnic Institute, London, N. With over 200 Illustrations. 8vo., 3s. 6d.
Manual
School Boards, Architects, Superintendents and Teachers. By RovvE, Ph.D. With 32 Illustrations. Crown 8vo., 31. ea". net.
Stuart H.
SL)L>OJV.B\JlLBER'S
TRADES.
Medium
8vo.
By
,
Col.
H. C. Seddon, R.E.
i6s,
With 25
Illustrations.
Crown
8vo.,
2.1.
6d.
giving the Cubic Contents of from One to Thirty Pieces Deals, Battens and Scantlings of the Sizes usually imported or used in the Building Trades, together with an Appendix showing a large number of sizes, the Contents of which may be found by referring to the aforesaid Tables. By Thomas Valder. Oblong 410., 6j. net.
Medium Svo. Part I. With 695 Illustrations, los. 6d. net. Part II. With 496 Illustrations, loj-. 6d. net. Part III. Materials. With 188 Illustrations, lis. net. Part IV. Calculations for Building Structures. With 551
Illustrations, 13J. net.
Scientific
Works published
by Longmans,
15
STEAM, OIL, AND GAS ENGINES. BALE.K HAND-BOOK FOR STEAM USERS being
;
Rules
for
Boiler
Engine Drivers and Boiler Attendants, with Notes on Steam Engine and Management and Steam Boiler Explosions. By M. Powis Bale, M.I.M.E., A.M.I. C.E. Fcp. 8vo., 25. hd.
By
Dugald
Member
of the Institution of Civil Engineers, Fellow of the Chemical of the Royal Institution, Fellow of the Institute of Patent
Illustrations.
Agents.
With 228
8vo., 15^.
With 27
Illustrations
FLETCHER.
M.Inst.Mech.E.
Medium
6,(.
6d. net.
Y^nGlASU
8vo.,
15.?.
By
George
With 212
C.
V.
Illustra-
An Instrument for measuring the Areas of Irregular Figures and specially useful for determining the Mean Effective Pressure from Indicator Diagrams from Steam, Gas and other Engines. Designed by D. A. Low, Professor of Engineering, East London Technical College, London. With Full Instructions for Use. is. net.
STEAM TURBINE.
Institute of
in the
By
8vo.
,
Robert
M.
etc.
Mechanical Engineers,
los. 6d. net.
Text.
With 207
8vo.
105. 6d.
B. Parsons, B.S., M.E., Consulting Engineer; American Society of Mechanical Engineers, American Society Engineers, etc. Professor of Steam Engineering, Rensselaer Poly
the
;
By H. DE
With 170
Illustrations.
8vo.
Crown
8vo.,
2s.
6d.
With 441
8vo., gs.
414 Diagrams.
8vo.
21s.
MANAGEMENT AND
Being a Treatise on Boiler Troubles and Repairs, Corrosion, Fuels, and Heat, on the properties of Iron and Steel, on Boiler By C. E. Stromeyer, Mechanics, Workshop Practices, and Boiler Design. Chief Engineer of the Manchester Steam Users' Association, Member of Council of the Institution of Naval Architects, etc. With 452 Diagrams, etc.
8vo., I2i. net.
Scientific
Co.
By SVANTE Arrhenius, Professor at the University of Stockholm. With 58 German Edition by John McCrae, Ph.D.
8vo.
,
<js.
bd. net.
:
C^i?^75-?J^/Z5(97\^. ELECTRO-DYNAMICS
Current Motor.
71
the DirectBy Charles Ashley Carus-Wilson, M.A. Cantab. With Diagrams, and a Series of Problems, with Answers. Crown Svc, js. td.
EXPERIMENCC/.(l/J//iV (P. ELECTRICITY TALLY. ByLiNN^UsCUMMiNG, M.A. With 242 Illustrations. Cr. 8vo. 4^. 6<i.
,
TREATED
<?Ci?^.THE
all
ART OF ELECTRO-METALLURGY,
and Diagrams.
8vo.,
\s.
Crown
8vo.
-js.
6d,
HIBBERT.
Diagrams.
MAGNETISM
By W,
Bvo., is,
Crown
AND
ITS
F.I.C,
ELEMENTARY
A.M.I.E.E.
MEASUREMENT.
Crown
Hibbert,
With
55
With 177
Illustrations.
Fcp. 8vo.,
3J. fxl,
/CKS-Ei?r. ELEMENTARY TREATISE ON ELECTRICITY AND MAGNETISM. By G. Carey Foster, F.R.S. and Alfred W.
;
B.Sc.
Illustrations
Founded on JOUBERT'S 'Traits El^mentair^ and Diagrams. 8vo. loj. 6if. net.
,
d'Electricit^
'.
/(?yC-E. EXAMPLES IN
By Samuel Joyce, A.I.E.E.
ELECTRICAL ENGINEERING.
Crown
8vo., ji.
^.i?i?/i^ffiZZ>. MAGNETISM AND DEVIATION OF THE COMPASS. By John Merrifield. LL.D., F.R.A.S., i8mo., 2j. td. /'^J?i?. PRACTICAL ELECTRICAL TESTING IN PHYSICS
By
8.f.
G. D.
With
231 Illustrations.
8vo.,
6if.
235 Illustrations.
2J,
6d.
ADVANCED ELECTRICITY AND MAGNETISM. With Crown 8vo. 317 bd. RHODES. K^ ELEMENTARY TREATISE ON ALTERIllustrations.
,
4f.
NATING CURRENTS.
Engineer.
(Vict.),
Consulting
SLINGO AND BROOXER.Works by W. SLINGO and A. BROOKER. ELECTRICAL ENGINEERING FOR ELECTRIC LIGHT ARTISANS AND STUDENTS. With 383 Illustrations. Crown 8vo. I2J. PROBLEMS AND SOLUTIONS IN ELEMENTARY
, ,
ELECTRICITY AND MAGNETISM. With 98 Illustrations. Cr. 8vo. 2j. rFiVZ?^ZZ. Worksby JOHNTYNDALL,D.C.L.,F.R.S. Seep 36.
Scientific
Works published
by Longmans,
Green,
Ss'
Co.
17
HOPKINS.
Institute,
PRO-
PERTIES.
PREECE
By Sir W. AND H. Preece, K.C.B., F.R.S., V.P.Inst., C.E., etc., Consulting Engineer and Electrician, Post Office Telegraphs and Sir J Sivewright, K.C. M.G. General Fcp. 8vo. 6s. iManager, South African Telegraphs. With 267 Illustrations.
;
. ,
S/FE WH/GNT.-TELEGRAPRY.
Description of Details
of Railway Construction subsequent to the completion of the Earthworks and Structures. By Sir John Wolfe Barry, K.C.B., F.R.S., M.I.C.E. With Fcp. 8vo., 4s. 6d. 218 Illustrations.
By John Goodman, Wh.Sch., M.I.C.E., M.I.M.E., With 714 Illustrations and numerous Examples. in the University of Leeds. Crown 8vo. 95. net.
GINEERS.
By David Allan Low London Technical College (People's Palace), London. With overiooo specially prepared Illustrations. Fcp. 8vo., gilt edges,
Professor of Engineering, East
js. 6d.
rounded corners,
loj.
6d.
net,
5J//7W. GRAPHICS,
or
tlie
Art
of Calculation
by Drawing
Lines apphed especially to Mechanical Engineering. By Robert H. Smith, With Part I. Professor of Engineering, Mason College, Birmingham. 8vo., 15J. separate Atlas of 29 Plates containing 97 Diagrams.
AND SIMILAR STRUCTURES; with Practical By BiNDON B. Stoney LL.D., Strength and other Properties of Materials. FRS MICE. With 5 Plates and 143 must, in the Text. Royal 8vo. 36^. UNWIN.i:YiE TESTING OF MATERIALS OF aCONSTRUCCollection of the
TION.
A Text-book for the Engineering Laboratory and By W. Cawthorne Unwin, I-.R.S., B.Sc. Results of Experiment. 8vo., i6j. net. Plates and 188 Illustrations ^nd Diagrams.
With
With
f-r^i?i?^7Y. ENGINEERING CONSTRUCTION IN STEEL AND TIMBER By William Henry Warren, Chalhs
University of Sydney. of Civil knd Mechanical Engineering, Royal 8vo. 161. net. Plates and 37s Diagrams.
,
IRON,
Professor
13 Folding
Drift,
Diagram.
Illustrations
and
Scientific
M.Inst.C.E.
With 368
NOTES ON DOCKS AND DOCK CONSTRUCTION. By C. COLSON, C.B., M.Inst.C.E. With 365 Illustrations. Medium 8vo., 21s. net CALCULATIONS IN HYDRAULIC ENGINEERING: a
Practical
Sanitary Engineering,
Text-Book
By
T.
Claxton Fidler,
I.
M.Inst.C.E.
Fluid Pressure and the Calculation of its Effects in EnWith numerous lUustns. and Examples. 8vo.,6.r. 6a'. net. Part II. Calculations in Hydro- Kinetics. With numerous IllusPart
gineering structures.
trations
and Examples.
8vo.
RAILWAY CONSTRUCTION.
trations
By W. H. Mills, M.I.C.E.,
With 516
Illus-
and Diagrams.
8vo.
i8j. net.
PRINCIPLES
trations.
By William Shield, F.R.S.E., M.Inst.C.E. Withgy IllusMedium 8vo. ly. net. TIDAL RIVERS: their (i) Hydraulics, (2) Improvement, (3) Navigation. By W. H. Wheeler, M.Inst.C.E. With 75 Illustrations. Medium 8vo. its. net.
, ,
N A V AITaRC HITECfu R E
WAR
Turning,
8vo.,
10s.
etc.,
of
War
Vessels.
Medium
bd.
net.
Crown
a Manual for Students of Science Classes and Draughtsmen Engaged and Naval Architects' Drawing Offices. With 114 Diagrams ^
:
NOLAfS.FRACTlCM. SHIPBUILDING
Treatise on
the
Struc ural Design and Building of Modern Steel \'e sels, the ^vork of construction, from the making of the raw material to the equipped vessel including subsequent upkeep and repairs. Bv A. Campbell Hoims, Member of thi Institution of Naval Architects, etc. In 2 vols. (Vol. I., Text, medium 8vo. vol. 11., Diagrams and Illustrations, oblong 4to.) 48f net
;
:
Iron Steel and Composite Vessels. By Thomas H. Watson, Lecturer on Naval Architecture at the Durham College of Science, Newcastle-upon-Tyne. With numerous Illustrations. Royal 8vo., ii;j. net.
A Manual
of Laying-
ATOJirffCOrT.-LATHES
Mecha-
338 Ilhostrations. 8vo.,,8.. '^ff/p7f^"v ^^7^%^\:^^L'X^ SffZZE V.-WORKSHOF APPLIANCES, including Descrip'^''''^'"^ Measuring Instruments, Hand-cutting Too^, 'p? La?hes nr'?)Li T"^ Lathes Drilling, Planemg, and other Machine Tools used by Engineirs Bv ^'''^'^ '"^ %" ^'^^"'^' Chapter on MllUng ty R R LsTE^^Wi'Jh' K. LISTER. ^"nf With 323 Illustrations. Fcp. Bvo., 55.
Scientific
Works published
by Longmans^
19
2.
A A
r/- net.
SQUARES. \i
I
2 45
rs
,, ,,
|
3.
Designs A & B. 45 to 60. 60 4" > 60 4" each i/- net. ( B i 6" VorJ B 2 60 6" ,, 1/3 ,, 60 2 60 8i"J 3 [33 60 84",, 2/- ,,
(Celluloid).
3.
Protractor No.
LOW'S ADJUSTABLE PROTRACTOR SET SQUARE, zs. 6d. net. LONGMANS' BLACKBOARD ENGLISH AND METRIC RULE. One
Metre
inches.
;
marked
2j. 6d.
in decimetres, centimetres,
inches, half-inches
and quarter-
**
ZC^. Works
ing,
by DAVID ALLAN LOW, Professor of EngineerEast London Technical College (People's Palace).
With 153
Illustrations
Crown
for measuring the Areas of Irregular Figures and specially useful for determining the Mean Effective Pressure from Indicator Diagi'ams from Steam, Gas and other Engine=. Designed by D. A. Low. With Full Instructions for Use. is. net.
An
Instrument
LOW AND
AND
DESIGN.
Low
M.I.Mech.E.
Bvo., 75.
General
Machinery.
Fastenings,
etc.
and
js. 6d.
Transmissive
Illustrations.
Fcp. 8vo.,
on Engine
Details.
With 259
BAUERMAN. Works
SYSTEMATIC
Fcp. 8vo.
,
DESCRIPTIVE
Fcp. 8vo,,
6s.
Illustrations.
BREARLEY
THE ANALYSIS OF and IBBOTSON. By HARRY Breakley and Fred STEEL-WORKS MATERIALS.
IBBOTSON, B.Sc. (Lond.
College, Sheffield.
),
With 85
By G.
With 56
Illustrations.
Fcp. 8vo.,
Oj.
Scientific
61^
Co.
E.TC.-Continiierl.
their Properties and Treatment. By A. K. Huntington, Professor of Metallurgy in King's College, London, and W. G. M'Millan, late Lecturer on Metallurgy in Mason's With 122 Illustrations. Fcp. 8vo., js. 6d. College, Birmingham.
British Isles,
A PRACTICAL TREATISE ON
With 209
Illustrations.
MINE SURVEYING.
L.
8vo.
j2s. net.
j?/ffi^Z>. METALLURGY.
tions.
By E.
Rhead, Lecturer on
With 94
Illustra-
35.
6d.
RHEAD
AND ^-^A'rOiV: ASSAYING AND GICAL ANALYSIS for the use of Students. Chemists and
Rhead,
,
METALLURAssayers.
By
E. L.
;
Lecturer on Metallurgy, Municipal School of Technology, Manchester and A. Humboldt Sexton, F.I.C, F.C.S., Professor of Metallurgy, Glasgow and West of Scotland Technical College. 8vo. loj. 61/. net.
By
,
F.
Rutlev, F.G.S.
With
4J. 6'/.
BALL. Works by Sir ROBERT S. BALL, LL.D., F.R.S. ELEMENTS OF ASTRONOMY. With 130 Figures and
grams.
Fcp. 8vo.
,
Dia-
6s.
6d.
A CLASS-BOOK OF ASTRONOMY.
Fcp. 8vo.
,
With 41
Diagrams.
IS.
6d.
Gill, F.R.A.S.
New
Edition
8vo.
,
arranged by
V. Merkifiei.D, li.A.
Medium
With
HERSCLLEL.O'UTl.mES OF ASTRONOMY.
F.
By
Sir
John
W. Herschel,
,
Bart.,
K.H.,
etc.
9 Plates
8vo.
I2J.
LA UGLLTON.AN
TICAL AND THEORETICAL STUDY OF NAUTICAL SURVEYING. By John Knox Laughton,M. A., F.R.A.S. With 35 Diagrams. Crown 8vo. 6j.
,
3/.4i?r/iV: NAVIGATION
Compiled by Staff Commander
Martin, R.N.
Royal 8vo.
iS.r.
Scientific
6r-
Co.
21
ASTRONOIWY, NAVIGATION,
the
E.TC.-Co>,linuea.
For
With
Students.
With Numerous
,
Examples and Examination Papers. By George W. Parker, M.A. Trinity College, Dublin. With 84 Diagrams. 8vo., s?. dd. net.
of
COMMON
TELE-
the Rev. T. W. Webb, M.A., F.R.A.S. Fifth Edition, Revised and greatly Enlarged by the Rev. T. E. EspiN, M.A., F.R.A.S. (Two Volumes.) Vol. I., with Portrait and a Reminiscence of the Author, 2 Plates, and numerous Illustrations. Crown 8vo., 6i. Vol. II., with numerous Illustrations.
Crown
8vo.,
6.t. (>d.
WORKS BY RICHARD
THE MOON
Condition.
: ,
A.
PROCTOR.
Her Motions, Aspect, Scenery, and Physical With many Plates and Charts, Wood Engravings, and 2 Lunar Photographs. Crown Svo. y. 6d.
the
Plurality
of
Worlds
With
14 Illustrations
Crown
Svo.
'^s.
6d.
INFINITIES
in
Abode
a Series of Essays conSpace and Time with the Infinities around us.
:
Crown
8vo.,
Familiar Essays
8vo., 3^. dd.
Natural Phenomena,
etc.
Crown
US;
Essays on the
Moon and
Crown
Planets,
y. 6d.
:
a Series of Essays on Suns Old, Young, and Dead. With other Science Gleanings. Two Essays on Whist, and Correspondence with Sir John Herschel. With 9 Star-Maps and Diagrams
3J.
6d,
Popular Guide
Instruction.
means
of
Amusement and
With
LIBRARY,
With an the Observatory, in Twelve Circular Maps (with Two Index-Plates). Introduction on the Study of the Stars. Illustrated by 9 Diagrams. Cr. Svo., 5^.
OVKR.
Scientific
WORKS BY RICHARD
PROCTOR 0/rf.
and Easy Guide
to
Plain
the
Showing in 12 Maps the position Constellations of the Southern Hemisphere. With an of the principal Star-Groups night after night throughout the year. True for every Year. Introduction and a separate Explanation of each Map.
Showing in r2 Maps the position of to the Knowledge of the Constellations. With Introthe principal Star-Groups night after night throughout the year. True for every Year. duction and a separate Explanation of each Map.
THE STARS
Knowledge
IN
THEIR SEASONS
Crown
8vo.
35.
6(/.
an
Easy Guide to a
Imperial 8vo.,
55-
of the Star-Groups.
In 12 Large Maps.
Familiar
Essays
on
Crown
8vo.,
8vo.,
y.
bd.
NATURE
STUDIES.
Clodd,
Crown
A.
y. 6d,
LEISURE READINGS.
Wilson, T. Foster, and
By R.
A. C.
Proctor, E. Clodd, A.
Crown
8vo.,
Ranyard.
y.
6d.
of the
and 247
Illustrations.
Crown
8vo.,
'zs.
6d.
ADVANCED GEOLOGY. A
Crown
8vo., js, 6d.
Manual
for
Students in Advanced
Classes and for General Readers. With over 300 Illustrations, a Geological Map of the British Isles (coloured), and a set of Questions for Examination.
With 236
Illus-
8vo., 2IJ.
MORGAN. WoxVs
Crown
8vo.
2j.
6d.
ADVANCED PHYSIOGRAPHY.
Crown
8vo., 4^.
With
215
Illustrations.
6d.
By T. Mellakd Reade F
8\o.,
2i.r.
F.RT.B.A., A.M.I.C.E,,
etc.
With
41 Plates.
net.
Scientific
Works published
by Longmans,
Green,
dr"
Co.
23
Crown Crown
ELEMENTARY PHYSIOGRAPHY:
an Introduction to the Study of Nature. With 13 Maps and 295 Illustrations. With Appendix on Astronomical Instruments and Measurements. Crown 8vo., 2J. dd.
ADVANCED PHYSIOGRAPHY.
Illustrations.
With
11
Crown
8vo.
4s.
6d.
NATURAL HISTORY AND GENERAL SCIENCE. FUJiN^AUX. Works by WILLIAM FURNEAUX, F.R.G.S.
THE OUTDOOR WORLD
book. Text.
,
AND STREAMS. With 8 Coloured Plates Crown Svo. BUTTERFLIES AND MOTHS (British). With 12 Coloured Crown Svo. the Text. Plates and 241 Illustrations THE SEA SHORE. With 8 Coloured Plates and 306 IllustraLIFE IN PONDS
and 331
Illustrations in the Text.
,
The Young
6s. net.
in
65. net.
Crown
Svo.
6s. net.
HUDSON
BRITISH
Works by W. H.
BIRDS.
Drawings by A. Thorburn, and 8 Plates and 100 Figures by C. and 3 Illustrations from Photographs. Crown Svo. 6s. net.
,
Lodge,
6s.
net.
F.Z.S.
(41 With 6 Photogravures and 66 from Drawings by the Author, Archibald Thorburn, and from Photo6s. net. Royal 4to., 6 graphs.
3 vols.
Quarto (13
in.
by 12
edgfs.
6 6s. net. Plates by the .\utfior and from Photographs. mrrica. *^* Only r,025 co/tifs friiih-J for liiigland and
1893-1896
Volume
4to.
With 44
Plates
Demy
Volume II. With 2 Charts and 17 Plates. Demy 4to., Volume III. With 33 Plates. Demy 4to., 32J. net. Volume IV. With 33 Plates. Demy 410. 2ri. net.
,
40J. net.
BIRDS.
Illustrations.
By E.
Crown
of Norwich.
With 160
24
Scientific
Co.
Series
Men and
Commerce
W.
J.
Crown
y'.6d.
net.'
By
F. T.
BROWN. \..\Y.O^\'\OV.\
STUDENTS:
Illustrations.
STUDIES
FOR
BREWING
a S>-stematic Course of Practical Work in the Scientific Principles underlying the Processes of Malting and Brewing. By Adrian T Bkou-x, M..Sc., Director of the School of Brewing, and Professor of the Biolo>iv and Chemistry of Fermentation in the Universitv of Birmingham, etc. With 36
8vo.
,
F.
CROSS and
E.
J.
BE VAN.
CELLULOSE
Crown
Elements of Plants. With reference to their Natural History and Industrial Uses. (C. F. Cross, E. J. Bevan- and C. Beadle.) With 14 Plates.
8vo., I2j. net.
RESEARCHES ON CELLULOSE,
6s. net.
1895-1900.
Crown
8vo.,
/ACA'SO.\:A TEXT-BOOK
TIONS:
with Examples. Bv W. lACKSox, A.R.C.S., Lecturer in Potterv and Porcelain Manufacture for the Staffordshire Education Committee and the Hanley Education Committee. Crown 8vo., 6d. net.
3.>-.
ON CERAMIC CALCULA-
MORRIS
by
AND WILKINSON. TU^ ELEMENTS OF COTTON SPINNING. By John Morris and F. Wilkinson. With a Preface
Crown
Sir B. A. Dobson, C.E., 8vo. js. 6d. net.
,
M.I.M.E.
Illustratiopb
By
in Brickwork and Masonry to the City and Guilds of London Institute, Head of Building Trades Department, Northern Polytechnic Institute, London, N. With over 200 Illustrations. Med. 8vo., 3^. 6d.
Bv
^^G^ov^F^^
8rown8^a
6.,
^^'''^"
^^^^'' ^^'^-
'''^''
^itb 20
Illustrations.
Scientific
Co.
25
^575^ K HEALTH IN
M.D,
.5
.
Ashby,
F.R.C.P.
Crown
8vo.,
y.
net.
C/CXrCiV: HEALTH IN
BuCKTON.
With
41
THE HOUSE.
Crown
By Mrs. C. M.
8vo.
,
2S.
By W. H. Cor-
dd.
by
J. L.
NOTTER,
M.A., M.D.,
HYGIENE. With 95 Illustrations. Crown 8vo., y. 6d. PRACTICAL DOMESTIC HYGIENE. With 83 Illustrations.
Crown
Bvo.,
2j.
6d.
M.D.
Illustrations.
THE DWELLING-HOUSE.
,
With 36
Illustrations.
Crown
3f. (>d.
With
11 Illustra-
Crown
THE EARTH
IN
being the Milroy Lectures delivered at the Royal College of Physicians in 1899, together with other Papers on Sanitation. With 13 Illustrations. Crown 8vo., 5J.
WILSON.K
Svo.
,
MANUAL OF
F.R.S.E., F.L.S.,
HEALTH-SCIENCE.
etc.
By
Crown
Andrew Wilson,
2j. 6d.
With 74
Illustrations.
26
Scientific
Works published
by
Longmans, Green,
(s'
Co.
PHYSIOLOGY,
etc.,
page
32.)
AS//B V AND
Physician to the General Hospital for Sick Children, Manchester and G. A. Wright, B.A., M.B. Oxon., F.R.C.S., Eng. Assistant-Surgeon to the ManEnlarged chester Royal Infirmary, and Surgeon to the Children's Hospital. and Improved Edition. With 192 Illustrations. 8vo. 25J.
,
By
With 68
BENNETT. Work's,
by Sir WILLIAM BENNETT, K.C.V.O., F.R.C.S., Surgeon to St. George's Hospital; Member of the Board of Examiners, Royal College of Surgeons of England.
CLINICAL LECTURES ON VARICOSE VEINS OF THE LOWER EXTREMITIES. With 3 Plates. 8vo., 6j. ON VARICOCELE A PRACTICAL TREATISE. With 4
;
8vo., 51.
CLINICAL
grams
in the
chiefly in relation to
With
12 Dia-
Text.
8^. 6d.
ON VARIX,
ITS
6d.
PASSIVE MOVEMENTS IN RECENT FRACTURES AND OTHER COMMON SURGICAL INJURIES The Treatment of Internal Derangements of the Knee Joint and Management
Illustrations.
of
Stiff Joints.
With
17
8vo.
6s.
CLINICAL EXAMINATION
By Richard
C.
Cabot, M.D., Physician to Out-patients, Massachusetts General Hospital. With 3 Coloured Plates and 28 Illustrations in the Te.xt. 8vo., i6j.
CARH,
PICA', RAN, and THE PRACTITIONER'S GUIDE. Bv Walter Carr, AI.D. (Lond.), F.R.C.P. T. Pickering Pick, F.R.C.S. Alban H. G. Doran, F.R.C.S. Andrew
I.
; ;
DUNCAN
M.R.C.P.
Duncan, M.D.,
B.Sc.
(Lend.),
F.R.C.S.,
8vo., 2ii.
net.
By Joseph
Professor of Pathology in the University of Glasgow. Fifth Edition. Revised throughout and Edited by Lewis R. Sutherland, M.D., Professor of Pathology, University of St. Andrews. With 729 Illustrations
and
Coloured Plates.
8vo.
aSj. net.
Scientific
Co.
27
niEDICINE
TREATMENT.
in
Part
The Treatment
Inflammation, Suppuration, Ulceration, Gangrene, Wounds and their Complications, Infective Dise.tses and Tumours the Administration of Anaesthetics. With 66 Illustrations. Royal 8vo. los. 6d.
,
Part II.
The Treatment
including the Skin and Subcutaneous Tissues, the Nails, the Lymphatic Vessels and Glands, the Fasciae, Bursae, Muscles, Tendons and Tendonsheaths, Nerves, Arteries and Veins. Deformities. With 141 Illustrations.
Royal 8vo.,
14-f.
Part III.
Part IV.
The Treatment
With 100
Amputations.
Illustrations.
of the Surgical Affections of the Joints (including Excisions) and the Spine. With 138 Illustrations, Royal 8vo. 141.
,
The Treatment
Part V.
of the Surgical Affections of the Head, Face, Jaws, Lips, Larnyx and Trachea; and the Intrinsic Diseases of the Nose, Ear and Larynx, by H. Lambert Lack, M.D. (Lond.), F.R.C.S., Surgeon to the Hospital for Diseases of the Throat, Golden Square, and to the Throat and Ear Department, The Children's Hospital, Paddington Royal 8vo., i8j. Green. With 145 Illustrations.
The Treatment
Part VI. Section I. The Treatment of the Surgical Affections of the Tongue and Floor of the Mouth, the Pharynx, Neck, Qisophagus, Stomach and Intestines. With 124 Illustrations. Royal Bvo. iSs.
,
Section
the
II.
The Treatment
Organs.
Rectum, Liver, Spleen, Pancreas, Throat, Breast and Genito-urinary With 113 Illustrations. Royal 8vo., 21J.
COOXE. Works
B.Sc, M.D.,
by
THOMAS COOKE,
Paris.
TABLETS OF ANATOMY.
Eleventh Edition tions given in the Westminster Hospital Medical School. in Three Parts, thoroughly brought up to date, and with over 700 Illustrations from all the best Sources, British and Foreign. Post 4to.
The Bones. 7^. 6d. net. 10s. 6d. net. Limbs, Abdomen, Pelvis. Part III. Head and Neck, Thorax, Brain. loj-. 6d. net. APHORISMS IN APPLIED ANATOMY AND OPERATIVE
Part
I.
Part II.
SURGERY.
etc.
Crown
6d.
By William
F.R.C.P., Obstetric Physician and Lecturer on With 394 Illustrations. Large George's Hospital, etc.
crown Bvo.
18^,
28
Scientific
DICKINSON. ^oiVi
Cantab., F.R.C.P.
Plates
by
W.
HOWSHIP DICKINSON,
Parts.
M.D.
12
With
4J.
(,d.
1855-
EJ?ICIfSN.THE SCIENCE
a Treatise on Surgical Injuries, Diseases, and Operations. By Sir John Eric Erichsen, Bart., F.R.S., LL.D. Edin., Hon. M.Ch. and F.R.C.S. Ireland. Illustrated by nearly looo Engravings on Wood. Royal 8vo., 48J. 2 vols.
FOWLER
LUNGS.
AND
GODLEE. THE
;
DISEASES
OF THE
By James Kingston Fowler, M.A., M.D., F.R.C.P., Physician to the Middlesex Hospital and to the Hospital for Consumption and Diseases of the Chest, Brompton, etc. and Rickman John Godlee, Honorary Surgeon in Ordinary to His Majesty, M.S., F.R.C.S., Fellow and Professor of Chnical Surgery, University College, London, etc. With i6o Illustrations. 8vo. 25J.
,
G^/?i?OZ). Works
M.D., F.R.S.,
by
Sir
etc.
A TREATISE ON
{14 Coloured),
(RHEUMATOID ARTHRITIS).
and 27
Illustrations
engraved on
Wood.
a Text-
Book of Oral Bacteria. By Kenneth W. Goadby, L.D.S. (Eng.). D.P.H. (Camb.), L.R.C.P., M.R.C.S., Bacteriologist and Lecturer on Bacteriology, National Dental Hospital, etc. With 82 Illustrations. 8vo., is. 6d. net.
By D. H. GooDSALL, F.R.C.S., Senior Surgeon, Metropolitan Hospital Senior Surgeon, St. Mark's Hospital and W. Eenest Miles, F.R.C.S., Assistant Surgeon to the Cancer Hospital, Surgeon (out-patients), to the Gordon Hospital, etc. (In Two Parts.) Parti. With 91 Illustrations.
8vo., 7i.
6(/.
net.
Henry Gray, F.R.S., late Lecturer on Anatomy at St. George's Hospital Medical School. The Fifteenth Edition Enlarged, edited by T. Pickering Pick, F.R.C.S., Consulting Surgeon to St. George's Hospital, etc., and by Robert Howden, M.A., M.B., CM., Professor of Anatomy in the University of Durham, etc. With 772 Illustrations, a large proportion of which are Coloured, the Arteries being coloured red, the Veins blue, and the Nerves yellow. The attachments of the muscles to the bones, in the section on Osteology, are also shown in coloured outline. Royal 8vo., 32J. net.
By
Scientific
Works published
IVIEDICINE
by Longmans,
29
HALLIBURTON. Works
PATHOLOGY.
With 104
AND
With 83
By
B.A., M.D., Secretary to the National Association for the Prevention of Consumption (England), Visiting Physician to the London Open-Air Sanatorium, Revised by Professor R. Koch. With 14 Illustrations.
Practitioners. By William Lang, F.R.C.S. Eng. Surgeon London Ophthalmic Hospital, Moorfields, etc. With 15 Crown 8vo., 3J, ^d.
,
FORENSIC
Crown
MEDICINE AND
P. Luff, M.D., B.Sc (Lend.), Physician Charge of Out-Patients and Lecturer on Medical Jurisprudence and Toxicology in St. Mary's Hospital. With 13 full-page Plates (i in colours) and
2 vols.
8vo., 24J.
/".^G^r. Edited by
PAGET.
STEPHEN PAGET.
By
Sir
JAMES
Bart.,
net.
By Charles D. F. Phillips, Inorganic Substances. late Lecturer on Materia Medica and Therapeutics Medical School late Examiner in the University Westminster Hospital at the 8vo., 2ij. of Edinburgh.
:
By W. H. and
is. td.
Fcap. 8vo-,
in Anaesthetics at
the
QVAIN.OyiAm'S
M.B., B.Ch.
,
(Sir
Richard)
DICTIONARY OF MEDI;
CINE. By Various Writers. Third Edition. Edited by H. Montague Murray, M.D., F.R.C.P., Joint Lecturer on Medicine, Charing Cross Medical assisted by John Harold, School, and Physician, Charing Cross Hospital
B.A.O.
,
St.
Elizabeth's Hospital;
and W. Cecil Bosanquet, M.A., M.D., M.R.C.P., Assistant Physician to With 21 Plates (14 in Colour) and numerous Charing Cross Hospital, etc. or 30^. net, half-morocco. 8vo., 21s. net, buckram Illustrations in the Text.
;
30
Scientific
Co.
ANATOMY.
F.R.S.
,
Professor
Dancer Thane,
Vol.
I.,
Part
I,
EMBRYOLOGY.
Vol.
III.,
Part
D.
II.
THE NERVES.
With
,
By
G.
Thane.
102
Illustrations.
Royal Bvo.
III.
9J.
Vol.
III.,
With 491
,
OF THE SENSES.
SchXfer, F.R.S.
trations.
Part
THE ORGANS
By E. A. With 178 Illus9J.
Illustrations.
Royal 8vo.
I.
12.J.
6i/.
Royal
8\'0.,
Vol.
II. ,
Part
OSTEOLOGY
Royal
II.
lUus.
8vo.,
lu.
Vol.
III.,
Part
,
IV.
SPLANCH-
Part
MYOLOGY
By G. D. Thane.
Roval 8vo.
ANGEIOLOGY.
With 199
1 6.1.
NOLOGY.
Illustrations.
Vol.
III.,
Part
I.
THE SPINAL
By E. A. With 139 Illus125. 6d.
Appendlx.
Royal
Bvo.,
SCHAFER. Works
Physiology
in the
Professor of
Descriptive
and
With 463
Illustrations.
8vo., 9j.net.
DIRECTIONS
PHY'SIOLOGY
:
IN
PRACTICAL
Elementary Physiology of Muscle and Nerve and of the Vascular and Nervous Systems. With 48 Diagrams and 24 pages of plain paper at end for Notes. Svo. net. 3
SMALE
AND CCZ I'.^'T?. DISEASES AND INJURIES OF THE TEETH, including Patholog)- and Treatment. By AIORTON Smale,
M.R.C.S.,
the School, Dental Hospital of
;
L.S.--\., L.D.S., Dental Surgeon to St. Mary's' Hospital, Dean of London, etc. and J. F. Colyer, L.R.C.P., M.R.C.S., L.D.S., Dental Surgeon to Charing Cross Hospital and to the Dental Hospital of London. Second Edition Revised and Enlarged by J. F. CoLYER. With 640 Illustrations. Large crown 8vo., lis. net.
SMITH {H.
iF:)._THE
B..\.,
By
Woodcuts.
Cr. 8vo., J.
Scientific
Works published
IVIEDICINE
by Longmans,
Green, Or Co.
31
of their Production and their Treatment. By Surgeon-General W. F. Stevenson, C.B. (Army Medical Staff), B.A., M.B., M.Ch. Dublin University, Professor of Military Surgery, Royal Army Medical College, London. With 127 Illustra8vo.
,
151. net.
TAPPEINER.
METHODS OF CLINICAL
INTRODUCTION
DIAGNOSIS.
TO
By
Dr.
CHEMICAL
H.
Tappeiner,
Professor of Pharmacology and Principal of the Pharmacological Institute of theUniversity of Munich. Translated by Edmond J. McWeeney, M.A., M.D. (Royal Univ. of Ireland), L.R.C.P.I., etc. Crown 8vo., 31. bd.
WALLER. ^oxY?.
by
AUGUSTUS
D.
WALLER,
M.D.,
Lecturer on Physiology at St. Mary's Hospital Medical School, London ; late External Examiner at the Victorian University.
Third
With 314
Illustrations.
8vo., iBj.
LECTURES ON PHYSIOLOGY.
Electricity.
First Series.
On Animal
8vo.
55. net.
FiTZWYGEAM,
Bart.
With 56 pages
of Illustrations.
Svo.
3^. net.
STEEL. \Noxks
A.V.D.,
by
late Professor of
Bombay
Veterinary College.
being
and Students.
With
Especially adapted for the use of Veterinary 88 Illustrations. Svo. ioj. 6d.
,
being a
,
Especially adapted for the use of Veterinary of Bovine Pathology. With 2 Plates and 117 Woodcuts. Bvo. 15J. Practitioners and Students.
being
a Manual of Ovine Pathology for the use of Veterinary Practitioners and Students. With Coloured Plate and 99 Woodcuts. Svo., I2i.
FOCMrr. Works
by
WILLIAM YOUATT.
7^.
THE HORSE. With 52 Wood Engravings. 8vo., THE DOG. With 33 Wood Engravings. 8vo., 6s.
6d
32
Scientific
PHYSIOLOGY, BIOLOGY, ZOOLOGY, ETC. (And see MEDICINE AND SURGERY, page 26.) ^5^^ K NOTES ON PHYSIOLOGY FOR THE USE OF
STUDENTS PREPARING FOR EXAMINATION.
M.D. Lond., F.R.C.P. Physician to Manchester. With 148 Illustrations.
,
By Henry Ashby,
i8mo.
y.
:
THE BODY
S.
By Mrs.
A. Barnett.
By
zs.
M.A. Oxon.
With 93
Illustrations.
Crown
8vo.
6d.
With 226
Illustrations.
With
BRODIE.
THE
ESSENTIALS
OF
,
EXPERIMENTAL
PHYSIOLOGY.
For the Use of Students. By T. G. Brodie, M.D., With 2 Lecturer on Physiolog}-, St. Thomas's Hospital Medical School. Plates and 177 Illustrations in the Text. 8vo. 6j-. 6d.
Introduction to the By Frederick Chapman, A.L.S., F.R.M.S. Study of the Protozoa. With 14 Plates and 42 Illustrations in the Text. 8vo. 9.^. net.
,
FURNEAUX.H\iM.K^ PHYSIOLOGY.
F.R.G.S.
By W. Furneaux,
6d.
With 218
Illustrations.
Crown
8vo.,
'zs.
HUDSON
ANIMACULES'.
Supplement
12s. 6d.
410., ioj. 6d. each. 4 Uncoloured Plates. Complete in 2 vols., with Supplement, 4to., ^ 45.
P. In 6 Parts.
or H. Gosse, F.R.S.
'
WHEEL-
MACALISTER. \WoT\is
M.D.
by
ALEXANDER MACALISTER,
With
41
105.
6d
With
u. 6d.
With 59
is.
6d,
MACDOUGAL.
DOUGAL,
With 159
tions.
\\oxV.% by DANIEL MACPh.D., Director of the Laboratories of the New York Botanical Garden.
TREMBLY
Svo.
75.
f>d.
net.
y.
Scientific
Works published
by Longmans,
Green, (r Co.
33
MORGAN.h^\m.K\. BIOLOGY
By
C.
an Elementary Text-Book.
With
Lloyd Morgan,
103 Illustrations.
Crown
8j. 6(/.
WORK
IN PRAC-
the Vascular and Nervous Systems. By E. A. Schafer, LL.D., F.R.S., Professor of Physiology in the University of Edinburgh. With 48 Diagrams.
y.
net.
THOHNTON. Works
by
JOHN THORNTON,
With
284
6s.
M.A. some
HUMAN
Coloured.
PHYSIOLOGY.
Crown
8vo.,
Illustrations.
Illustrations,
ELEMENTARY BIOLOGY,
With numerous
Illustrations (6 of
Experimental.
With
6d.
178
Crown
8vo.
3^.
BACTERIOLOGY.
CURTIS.TYl^ ESSENTIALS OF PRACTICAL BACTERIOLOGY An Elementary Laboratory Book for
:
By
H.
J.
&\o.,gs.
ELEMENTARY BACTERIOLOGY.
,
By M. L.
University of
CM.
Edin.
Diplomate
in State
Medicine,
With Coloured
Frontispiece
and 26
Crown
8vo.
3J. net.
^iP^ira'Z^iVZ'. MICRO-ORGANISMS
IN WATER.
To-
gether with an Account of the Bacteriological Methods involved in their Specially designed for the use of those connected with the Investigation. By Percy Frankland, Ph.D., B.Sc. Sanitary Aspects of Water-Supply. (Lond.), F.R.S., and Mrs. Percy Frankland. With 2 Plates and Numerous
Diagrams.
8vo.
165. net.
FRANKLAND. BACT^V^A
Percy Frankland, F.R.M.S.
Book of Oral Bacteriologist
IN DAILY LIFE.
Crown
8vo.
,
By Mrs.
Text;
5^.
net.
Bacteria. By Kenneth W. Goadby, L.D.S. Eng., etc, and Lecturer on Bacteriology, National Dental Hospital, etc.
in the University of
a Laboratory G. E. Allan, B.Sc, Lecturer Birmingham, and J. H. Millar, F.I.C, formerly Lecturer School of Malting and Brewing, and Revised by the Author.
: ,
12.^.
6j.
net.
34
Scientific
^/rTT^TV. ELEMENTARY
By Edith Aitken, Crown 8vo. 4J. td.
,
late
BENNETT
on Botany
and iJ/67?^^
K HANDBOOK
;
OF CRYPTOF.L.S., Keeper
GAMIC BOTANY.
and George at St. Thomas's Hospital With 378 Illustrations. of Botany, British Museum.
Murray,
8vo., 1&5.
CJiOSS AND
and C.
BE VAN.Works BEADLE.
:
by C. F.
CROSS,
E. J.
BEVAN
Industrial
CELLULOSE
Uses.
Elements of Plants.
With 14
and
EDMONDS.Works
8vo., IS. 6d.
RESEARCHES ON CELLULOSE, 1895-1900. Cr.8vo.,6.f.net. by HENRY EDMONDS, B.Sc, London. ELEMENTARY BOTANY With 341 Illustrations. Cr.8vo.,2j.6^. BOTANY FOR BEGINNERS. With 85 Illustrations. Fcp.
STUDY OF BOTANY
With
121 Illustrations.
Crown
HOFJ^MAJVJV.Al.VmE FLORA
Botanists.
for Tourists and Amateur By Dr. JULius Hoffmann. Translated by E. S. Barton (Mrs. A Gepp). With 40 Plates, containing 250 Coloured Figures, from Water-Colour Sketches by Hermann Friese. With Text descriptive of the most widely dis:
tributed
8vo.
Adapted
to
Home
With 20
and
Illustrations.
Cr. 8vo.,5J.
LINDLEY
Edited by
Plates
AND
J.
Lindley, M.D., F.R.S., and T. Moore, F.L.S. and numerous Woodcuts. Two parts. Fcp. 8vo., i2j.
J/ciV:^^. CLASS-BOOK
OF BOTANY.
,
MORPHOLOGY
OLOGY.
Fcp. 8vo.
,
AND
With
42
bd.
PHYSIDiagrams.
CLASSIFICATION
ij.
By W. R. McNab. OF PLANTS.
Fcp. 8vo.,
li.
Translated by F. E. Weiss,
With 33
THOME
AND PHYSIO-
Scientific
Works published
by Longmans,
Green,
(s^
Co.
35
^OTMiX Continued.
TU-B-E
CRYPTOGAMIC
Introduction to the Study of Pathogenic Fungi, Slime Fungi, Bacteria and Algae. By Dr. Karl Freiherr von TUBEUF, Privatdocent in the University of Munich. English Edition by William G. Smith, B.Sc, Ph.D., Lecturer on Plant Physiology, University Edinburgh. 8vo. of With 330 Illustrations. Royal i8j. net.
,
UK DISEASES OF PARASITES.
PLANTS
INDUCED
BY
For the use of Elementary By W. Marshall Watts, D.Sc. Lond. Cr, 8vo., 2s. 6d.
With 49
Illustrations.
Crown
COLEMAN
Frank
Svo.,
ij.
AND ADDYMAN. V^ACTICAE AGRICULTURAL CHEMISTRY. By J. Bernard Coleman, A.R.C.Sc, F.I.C, and
T.
Addyman,
With 24
Illustrations.
Crown
6d. net,
HAGGARD. Works
With 36
Illustrations
yi. 6d, net.
by H.
:
RIDER HAGGARD.
A FARMER'S YEAR
being his Commonplace Book for 1898. Crown Svo. by G. Leon Little and three others.
:
RURAL ENGLAND
Maps and
75 Illustrations
/EJS YLL. Works by GERTRUDE JEKYLL. HOME AND GARDEN Notes and Thoughts,
:
Practical
and
Svo.,
Critical, of
Worker
in both.
With 53
Illustrations
from Photographs.
a Working Amateur.
WEATHERS.
PLANTS.
of the Hardiest and most Beautiful Annuals and Biennials, Hardy Herbaceous and Bulbous Perennials, Hardy Plants, Flowering Ornamental Water and Bog and Trees and Shrubs, Conifers, Hardy Ferns, Hardy Bamboos and other Ornamental Grasses and also the best icinds of Fruit and Vegetables that may be grown in the Open Air in the British Islands, with Full and Practical Instructions as to Culture and Propagation. By John Weathers, F.R.H.S., late Assistant Secretary to the Royal With 163 Horticultiural Society, formerly of the Royal Gardens, Kew, etc. Diagrams. 8vo., 21J. net.
Containing Descriptions
IVBB. Works
With 34
Illustrations.
Crown
for
Svo.
2s. 6d.
AGRICULTURE. A
With 100
Illustrations.
Manual
Crown
^6
Scientific
Co.
WORKS BY JOHN TYNDALL, D.C.L., LL.D., F.R.S. LECTURES ON SOUND. With Frontispiece of Fog-Syren, and
203 other Woodcuts and Diagrams in the Text.
Crown
8vo.
ioj. hd.
HEAT, A
MODE OF MOTION.
Crown
1872
8vo., izj.
and
Diagrams.
Crown
AND
With
Portrait,
FRAGMENTS OF SCIENCE:
Addresses, and Reviews.
Vol.
I.
Vol.
II.
The Constitution of Nature Radiation On Radiant Heat in Relation to the Colour and Chemical Constitution of Bodies New Chemical Reactions produced by Light On Dust and Disease Voyage to Algeria to observe the Eclipse Niagara The Parallel Roads of Glen Roy Alpine Sculpture Recent Experiments on FogSignals On the Study of Physics On Crystalline and Slaty Cleavage On Paramagnetic and Diamagnetic Forces Physical Basis of Solar Chemistry Elementary Magnetism On Force Contributions to Molecular Physics Life and Letters of Faraday The Copley Medallist of 187c The Copley Medallist of 1871 Death by Lightning Science and the Spirits. Reflections on Prayer and Natural Law Miracles and Special Providences On Prayer as a Form of Physical Energy Vitality Matter and Force Scientific Materialism An Address to Students Scientific Use of the Imagination The Belfast Address Apology tor the Belfast Address The Rev. James Martineau and the Belfast Address Fermentation, and Bearings on Surgery and Medicine Spontaneous Generation Science and Man Professor Virchow and Evolution The
its
2 vols.
Essays,
Electric Light.
Contents. The Sabbath Goethe's Farbenlehre Atoms, Molecules, and Ether Waves Count Rumford Louis Pasteur, his Life and Labours The Rainbow and Congeners Address delivered at the Birkbeck Institution on October 22, 1884 Thomas Young Life in the Alps About Common Water Personal Recollections of Thomas Carlyle On Unveiling the Statue of Thomas Carlyle On the Origin, Propagation, and Prevention of Phthisis Old Alpine Jottings A Morning on Alp Lusgen. ESSAYS ON THE FLOATING MATTER OF THE AIR IN RELATION TO PUTREFACTION AND INFECTION. With 24 Wood' '
NEW FRAGMENTS.
Crown
8vo., lo^.
(,d.
its
cuts.
Crown
8vo.
']s.
^d.
Crown
Crown
Bvo.
is.
6d.
Crown
8vo.
is.
6d.
With
58
Crown
Bvo., at.
6rf.
being a Narrative of Excursions and Ascents. An Account of the Origin and Phenomena of Glaciers, and an Exposition of the Physical Principles to which they are related. With Crown 8vo. 6s. 6d. net. 7 Illustrations.
HOURS OF EXERCISE
Crown
8vo., 6j. 6d. net.
IN
THE
ALPS.
Crown
With
7 Illustrations.
FARADAY AS A DISCOVERER.
Scientific
(s'
Co.
37
PHOTOGRAPHY.
8vo.,
TEXT-BOOKS OF SCIENCE.
By
Sir
With 66
3J.
Illus-
Fcp. 8vo.,
dd.
RAILWAY APPLIANCES.
Wolfe Barry,
M.I.C.E.
Fcp. 8vo.
,
By Sir John
F.R.S.,
Illustrations.
K.C.B.,
With 218
4^. td.
6d.
By George
C.B., Chairman of the Board of Works, Dublin. With 212 Illustrations. Fcp. 8vo. 6s.
3J.
6d.
By T.
E.
Thorpe, C.B.,
QUALITATIVE
T. E.
ANALYSIS
LABORATORY PRACTICE.
H. Preece, K.C.B., F.R.S., M.I.C.E., and Sir J. Sivewright, M.A., K.C.M.G. With
Fcp. 8vo.,
6s.
and M. M. Pattison Muir, M.A. and F. R. S. E. With Plate of Spectra and 57 Illustrations. Fcp. 8vo. y. 6d.
,
PHYSICAL
INTRODUCTION TO THE STUDY TECHNICAL ARITHMETIC AND OF CHEMICAL PHILOSOPHY. MENSURATION. By Charles By William A. Tilden, D.Sc,
London, F.R.S. With Illustrations. Fcp. 8vo., 5J. With Answers to Problems. Fcp. 8vo. y. 6d.
,
W. Merriefield,
Fcp. 8vo.,
M.A.
ELEMENTS OF ASTRONOMY.
, ,
By
By Frank
,
Sir R. S. Ball, LL. D. F. R.S. With 130 Illustrations. Fcp. Bvo. 6s. 6d.
RUTLEY, F.G.S. With 6 Plates and 88 Illustrations Fcp. 8vo. 4J. 6d.
WORKSHOP APPLIANCES,
Descriptions of
including
SYSTEMATIC
373
MINERALOGY.
Fcp. 8vo,,
6s.
By With By
etc.
,
some
of the
Machine
DESCRIPTIVE MINERALOGY.
Hilary Bai^rman,
With 236
Illustrations.
Fcp. 8vo.
55.
6s.
Part
ings
I.
With 345
7J. 6d.
Illustrations.
Fcp.
8vo.,
6d.
Part
By
J.
II.
Chiefly
on Engine DeIllustrations.
THEORY OF HEAT.
Maxwell,
M.A.,
F.R.SS., L. & E. tions. Fcp. 8vo.,
Clerk
tails.
With
,
259
Fcp. Bvo.
6s.
41. 6d.
PRACTICAL PHYSICS.
tions.
By R. T. Glazebrook, M.A., F.R.S., and W. With N. Shaw, M.A. 134 IllustraFcp. 8vo.,
7s. 6d.
By
Fcp.8vo.,3J.6i/.
38
Scientific
Works
pi4blished by
By
GEOLOGY
and
an
Appendix
of
Examination
8vo.
,
Questions.
Crown
4J. ^d,
Readers. By Charles Bird, B.A. (Lond.), F.G.S. With over 300 Illustrations, a Geological Map of the British Isles {coloured), and a set of Questions for Examination. Crown
8vo., js. 6d.
THEORETICAL MECHANICS.
Solids, including Kinematics, Statics, and Kinetics. Bv A. Thornton,
HUMAN PHYSIOLOGY
^
M.A., F.R.A.S.
tions,
130 over 900 Examples from Examination Crown 8vo. \s. Gd. Papers, etc.
,
a Manual for Students in advanced Classes of the By Science and Art Department. John Thornton, M.A. With 284
:
Illustrations,
some
of
which
,
are
HEAT.
By
Mark
R.
Wright, Hon.
With 136 Illustrations and numerous Examples and Examination Papers. Crown 8vo.
Inter. B.Sc. (Lond.).
4J.
PHYSIOGRAPHY.
Deposits.
By John Thorn8vo.,
ton, M.A. With II Maps, 255 Illustrations, and Coloured Map of Ocean
6d.
Crown
4.?.
6d.
LIGHT.
AGRICULTURE.
Ph.D., B.Sc.
Crown
M.D.
8vo.
js.
By J. LaneNottek, M.A.,
Illustrations.
Crown
8vo.
6d.
Army
late
m the Army Medical School, Netley, Major, Royal Army Medical Corps.
With 95
3J.
Illustrations.
Crown
8vo.,
Crown
8vo.
4J. 6d.
6d.
Illus-
GEOMETRICAL DRAWING FOR BUILDING CONSTRUCTION AND ART STUDENTS. Embracing DRAWING. By Edward J.
Plane Geometry and
the
its
Applications,
Use
of Scales,
2s.
Elevations of Solids.
Crown
-
6d.
Svo.
TEXT BOOK ON PRACTICAL, AN ELEMENTARY COURSE OF SOLID, OR DESCRIPTIVE GEOM.\THEM.\TICS. Containing .ArithMETRY.
Svo.
,
(Whitworth Scholar).
2S.
Part
metic
tions
Euclid (Book
E-xercises)
,
I.,
;
with Deduc-Algebra.
and
Svo.
and
Crown
zs.
6d
Scientific
Co.
39
EL.EIVIENTARY SCIENCE
MiMMJ^- Continued.
PHYSIOGRAPHY.
Instru-
THEORETICAL MECHANICS.
,
In-
ELEMENTARY
By
By J. E. Taylor, M.A. B.Sc. With numerous Examples and Answers, and 17s Diagrams and Illustrations. Crown 8vo. 2.;. 6/.
,
Crown
Fatty
2J. 6d.
ORGANIC CHEMISTRY:
Compounds.
,
the
By
J.
E.
Taylor, M.A.,
163 Illustrations, over 500
^d.
B.Sc. (Lond.).
With
By R. Lloyd Whiteley, F.I.C., F.C.S. With 45 Illustrations. Crown 8vo. 3J. 6d.
THEORETICAL
FLUIDS.
By
J.
MECHANICSE.
Taylor, M.A.
122 Illustrations,
2J-.
6d.
B.Sc. (Lond.).
With
numerous Worked Examples, and about 500 Examples from Examination Papers, etc.
AN INTRODUCTION TO
By William Jago, Crown 8vo., is. 6d.
PRACTI-
Crown
8vo.
-zs.
6d.
MANUAL OF MECHANICS.
GOODEVE, M.A.
With
PRACTICAL
William
3vo.
,
CHEMISTRY
A. Tilden, D.Sc.
the
138 Illustrations and Diagrams, and 188 Examples taken from Examination Papers, with Answers. By T. M.
By
Fcp.
6d.
Crown
8vo.
2j. 6rf.
F.R.G.S. Crown 8vo., 2i. 6rf. SOUND, LIGHT, AND HEAT. By Mark R. Wright, M.A. With 160 ELEMENTARY GEOLOGY. Diagrams and Illustrations. Crown
8vo.
,
2J. 6d.
B.A., F.G.S. With Geological Map of the British Islands, and 247 Illustrations.
Charles Bird,
Coloured
METALLURGY:
Book.
an Elementary Text-
Crown
By
E. L.
Illustrations.
With 94
,
3^. 6d.
HUMAN PHYSIOLOGY.
William
With 218
2s. 6d.
By
8vo.
Furneaux,
Illustrations.
F.R.G.S.
Crown
PHYSICS.
Course. By Wright, M.A. With 242 A Crown 8vo. 2j. (>d. Illustrations.
Alternative
Mark
R.
MENTARY BIOLOGY.
BiDGOOD, B.Sc.
By
J.
With 226
Illustra-
tions.
Crown
W.
POYSEE, M.A.
With 235
,
Illustrations.
Crown
8vo.
2J.
6d.
With"*
,
2s. 6d.
bracing a Complete Set of Answers to the South Kensington Papers for the years 1885-1899, and a Series of Original Questions. With 67 OrigiBal Illustrations.
With 185
AGRICULTURE.
Ph.D.
8vo.
,
By Henry
J.
Webb,
Crovyi)
With 34
Illustrations.
Crown
8vo.
25.
2S. (id.
40
Scientific
ds'
Co.
Carey Foster,
F.R.S., and by Sir Philip Magnus, B.Sc, B.A., of the City and Guilds of London Institute.
|
ASTRONOMY.
4T
By Sir Robert Sta- ZOOLOGY OF With ANIMALS. WELL Ball, LL.D., F.R.S
Dia^ams.
,
THE VERTEBRATE
By ALEXANDER MacWith 59 Diagrams.
6d.
Fcp. 8vo.
ij.
6rf.
alister, M.D.
Fcp. 8vo.
,
MECHANICS. By
89 Diagrams.
Sir
Robert StaWith
(>d.
,
IJ.
HYDROSTATICS
TICS.
B.Sc,
Fcp. 8vo.
AND PNEUMAPhilip
M A., M.D., F.R.C.P. With 22 Illustrations. Fcp. 8vo. u. PHYSICS AND MOLECULAR
,
By W.
(sd.
By
B.A.
,
Sir 6d.
Magnus,
With 79
2s.)
IS.
witk Answers,
The Worked
2S.
With 91 Diagrams.
Fcp.
8vo.
\s. (jd.
GEOMETRY,
URES.
F.R.S.
8vo.
,
BOTANY.
,
IS.
6d.
With 118 Diagrams. Fcp.Svo.,is.6d. Outlines of Morphology and Physiology. By W. R. McNab, With 42 Diagrams. M.D. Fcp.
By Richard
,
Macalister, M.D.
grams.
Fcp. 8vo.
,
WoRMELL, M.A.,
Diagrams.
i.r.
By Joseph
Illustrations.
S.
With 152
trum.
Crown
2J.
6d.
I.
By
A.
G.
Cracknell, M.A.,
8vo.
,
B.Sc.
Crown
3J.
6d.
M.A
2J.
S.
108
Crown
8vo.
6d.
6d.
SICS.
Stage
I.
By W. Watson,
,
With 83 Illustrations.
Cr. 8vo.
2j.
6d.
F.R.S., D.Sc. With 120 Illustrations and 193 Exercises. Cr. 8vo. zj. 6d.
By John
108 Illus6d.
By
I.
Bretland
With
,
Farmer,
F.R.S., M.A.
121
6d.
Illustrations.
ELEMENTARY
GIENE.
S.
Crown
I.
8vo.
2J.
PRACTICAL
HYto
Section
By William
Furneaux.
With .Appendix
ELEMENTARY PRACTICAL BUILD- ELEMENTARY PRACTICAL PHYING Cr.JNSTRUCTION. Bv F. W. Booker. With Illustrations.
Crown
ff,
3J.
6d.
SIOLOGY.
M.A.
8vo.
2j. 6i^,
^j.
6d,
000/10/04.