GEOMETRY AND INTUITION

Author(s): Hans Hahn

Source: Scientific American , Vol. 190, No. 4 (April 1954), pp. 84-91
Published by: Scientific American, a division of Nature America, Inc.
Stable URL: https://www.jstor.org/stable/10.2307/24944527

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Scientific American, a division of Nature America, Inc. is collaborating with JSTOR to digitize,
preserve and extend access to Scientific American

This content downloaded from

86.60.213.150 on Wed, 29 Sep 2021 14:44:16 UTC
All use subject to https://about.jstor.org/terms
 GEOMETRY AND INTUITION
A classic dp,scription of ho,v "common sense," once

accepted as the basis of physics but no,v rejected,

is also inadequate as a foundation for mathematics

We have grown so accustomed to the revolutionary nature Some years ago the brilliant Austrian mathematician Hans
of modern science that any theory which affronts common Hahn surveyed the situation in a Vienna Circle lecture which
sense is apt to be regarded today as half proved by that very he titled "The Crisis in Intuition." His analysis is still fresh
fact. In the language of science and philosophy the word and timely, and it is published here, in part, for the first time
for common sense is intuition-it relates to that which is in English. Hahn began with Immanuel Kant, the foremost
directly sensed or apprehended. Twentieth-century discov­ exponent of the importance of intuition, and showed how the
eries have dealt harshly with our intuitive beliefs about foundations of Kant's ideas about knowledge "have been
the physical world. The one area that is commonly sup­ shaken" by modern science. The intuitive conceptions of
posed to remain a stronghold of intuition is mathematics. space and time were jolted by Einstein's theory of relativity
The Pythagorean theorem is still in pretty good shape; the and by advances in physics which proved that the location of
self-evident truths of mathematics are in the main still true. an event in space and time cannot be determined with unlim­
Yet the fact is that even in mathematics intuition has been ited precision. Hahn went on to consider the demolition of
taking a beating. Cornered by paradoxes-logical contradic­ Kant's ideas about mathematics, and he illustrated his theme
tions-arising from old intuitive concepts, modern mathe­ with the case of geometry, where "intuition was gradually
maticians have been forced to reform their thinking and to brought into disrepute and finally was completely banished."
step out on the uncertain footing of radically new premises. This section of his lecture, somewhat condensed, follows.

by Hans Hahn

ne of the outstanding events in for example, of a wave curve such as is Weierstrass invented the curve by an

O [the banishment of intuition from

geometry] was the discovery that,
in apparent contradiction to what had
shown here [Figure 1]. In the neighbor­
hood of the point p the curve has infinite­
ly many waves. The wavelength and the
intricate and arduous calculation, which
I shall not attempt to reproduce. But his
result can today be achieved in a much
previously been accepted as intuitively amplitude of the waves decrease without simpler way, and this I shall attempt to
certain, there are curves that possess no limit as they approach p. If we take suc­ explain, at least in outline.
tangent at any pOint, or-what amounts cessive points closer and closer to p, the We start with a simple figure which
to the same thing-that it is possible to average slope between p and p' (the consists of an ascending and a descend­
imagine a point moving in such a man­ moving point) drops from plus 1 through ing line [Figure 2]. We then replace the
ner that at no instant does it have a defi­ o to minus 1 and then rises from minus 1 ascending line with a broken line in six
nite velocity. The questions involved to plus 1. That is, as p' approaches p parts, first rising to half the height of the
here directly affect the foundations of without limit through infinitely many
the differential calculus as developed by waves, the average slope between p and
Newton and Leibnitz. '
p keeps oscillating between tl e values 1
Newton calculated the velocity of a and -1. Thus there can be no question of
moving point at the instant t as the limit­ a limit or of a definite slope of the curve
ing value approached by the average at the pOint p. In other words, the curve
velocity between t and an instant close we are considering has no tangent at p.
to it, t', as t' approaches t without limit. This relatively simple illustration
Leibnitz similarly declared that the demonstrates that a curve does not have
slope of a curve at a point p is the lim­ to have a tangent at every point. Never­
iting value approached by the average theless it used to be thought, intuitively,
slope between p and a nearby point p' that such a deficiency could occur only
as p' approaches p without limit. at exceptional points of a curve. It was
Now one asks: Is this true for every therefore a great surprise when the great
curve? It is indeed for all the old familiar Berlin mathematician Karl Weierstrass
ones-circles, ellipses, hyperbolas, parab­ announced in 1861 a curve that lacked
olas, cycloids, etc. But it is not true, a precise slope or tangent at any pOint. Figure 1

84

© 1954 SCIENTIFIC AMERICAN, INC

This content downloaded from

86.60.213.150 on Wed, 29 Sep 2021 14:44:16 UTC
All use subject to https://about.jstor.org/terms
original line, then dropping all the way it will pass through all the points of a crease. It is not difficult to prove that this
down, then again rising to half the slluare-and yet no one would consider figure, in spite of its linear character,
height, continuing to full height, drop­ the entire area of a square as simply a cannot be generated by the motion of a
ping back again to half height and finally curve. With the aid of a few diagrams point, for no motion of a point is conceiv­
rising once more to full height [Figllre I shall attempt to give at least a general able that would carry it through all the
.s J. We replace the descending line also idea of how this space-filling motion is points of this wave curve in a finite time.
with a broken line of six similar parts. generated.
From this figure of 12 line segments we Divide a square into four small wo important lluestions now suggest
evolve, again by replacing each segment squares of equal size and join the center T themselves. (1) Since the time­
with a broken line of six parts, a figure of pOints of these squares by a continuous honored definition of a curve fails to
72 line segmeGts [Figure 4J. It is easy to curve composed of straight-line seg­ cover the fundamental concept, what
see that repetition of this procedure will ments [Figure 5J. Now imagine a point serviceable definition can be substituted
lead to more and more complicated fig­ moving at uniform velocity so that it will for it? (2) Since the class of geometric
ures. It can be demonstrated that the traverse the continuous curve made of objects that can be produced by the mo­
geometric objects constructed according these line segments in a certain unit of tion of a point does not coincide with the
to this rule approach without limit a time. Next divide each of the four class of all curves, how shall the former
definite curve possessing the desired squares again into four equal squares so class be defined? Today both questions
property; namely, at no point will it that there are 16 squares, and connect are satisfactorily answered; I shall defer
have a precise slope, and hence at no their center points [Figure 6J. Imagine for a moment the answer to the first
point a tangent. The character of this the point moving so that in the same time question and speak briefly about the
curve of course entirely eludes intuition; as before it will traverse this second second. This was solved with the aid of
indeed, after a few repetitions of the seg­ curve at uniform velocity. Repeat the a new geometric concept-"connectivity
menting process the evolving figure has procedure, each time imagining the in the small." Consider a line, a circle or
grown so intricate that intuition can pOint to move so that in the same unit of a square. In each of these cases, we can
scarcely follow. The fact is that only time it will traverse the new system of move from one pOint on the figure to
logical analysis can pursue this strat�ge lines at a uniform velocity. Figure 7 another very close to it along a path
object to its final form. shows one of the later stages, when the which does not leave the confines of the
original square has been divided into figure, and we remain always in close
ISst it be supposed that intuition fails 4,096 small squares. It is now possible proximity to both points. This is the
only in the more complex branches to give a rigorous proof that the succes­ property called "connectivity in the
of mathematics, I propose now to exam­ sive motions considered here approach small." Now the wave curve we have
ine a failure in the elementary branches. without limit a curve that takes the mov­ just considered does not have this prop­
At the very threshold of geometry lies ing point through all the pOints of the erty. Take for example the neighboring
the concept of the curve; everyone be­ large square in the given time. This mo­ points p and q [Figure 9J. In order to
lieves that he has an intuitively clear tion cannot possibly be grasped by in­ move from p to q without leaving the
notion of what a curve is. Since ancient tuition; it can only be understood by curve it is necessary to traverse the in­
times it has been held that this idea logical analysis. finitely many waves lying between them.
could be expressed by the following While a geometric object sLlch as a The points on this path are not all in
definition: Curves are geometric figures square, which no one regards as a curve, close proximity to p and q, for the waves
generated by the motion of a point. But can be generated by the motion of a all have the same amplitude.
attend! In the year 1890 the Italian point, other objects which one would not It is important to realize that "con­
mathematician Giuseppe Peano (who is hesitate to classify as curves cannot be so nectivity in the small" is the basic char­
also renowned for his investigations in generated. Observe, for instance, the acteristic of figures that can be generated
logic) proved that the geometric figures wave curve shown here [Figure 8J. In by the motion of a point. A line, a circle
that can be generated by a moving point the neighborhood of the line segment ab and a square can be generated by the
also include entim plane surfaces. For the curve consists of infinitely many motion of a point because they are con­
instance, it is possible to imagine a point waves whose lengths decrease without nected in the small; the wave figure
moving in such a way that in a finite time limit but whose amplitudes do not de- shown cannot be generated by the mo-

Figure 2 Figure 3 Figure 4

85

© 1954 SCIENTIFIC AMERICAN, INC

This content downloaded from

86.60.213.150 on Wed, 29 Sep 2021 14:44:16 UTC
All use subject to https://about.jstor.org/terms
 
ċ‹ċ 
ċŽċ

ſ $ſ ſ .
ſ   ſ 
ſ  ſ 
ſ ;  & ' Zc K$
ſ
 ſ  ſ ſ Lſ

ſſ
ſ Ð5ſ  
ſ [ſ   ſ

ſ 
ſ
ſ ſ
ſ
ſſ   ſ    ſ$ſ
ſ ;  ſ ſ .  ſ
ſ  ſ 
ſ
ſ;
.
ſ $ſ 

Lſ  ſ  ſ ĩż 1ſ  .ſ

ſſ ſ
1N ſ ſ 
ſ1 
ſa  ;

ſ   ſ
ſDN$ſ ſ$ſſ
ſ

 Lſ 
ſſ  ſ9 .5ſ_¥ſ$ſNſ  Mſ .
ſ 
ſ  ſ 
ſ
 ſ
.ſ$ſ
ſX1ſ  
 ſ  $ ' 
ſ$ſ
ſ
ſ
ſ  
 ſ  % '
5 _ſſ 
ſ.
ſ
ſ ſ Nſ 5 _ Mſ  
ſ Pſ   ſ

ſ

ſ  
 ſ  ſ
1
R ¯ ſ 
ſ 
ſ 1ſ <ſ ſ 
ſ  ſ 9:
n
¯  
ſ  Uſ €.
ſ Cm Nſ
 ſ ſ ſ
ſ
ſ

ſ  ſ
ſ
C6'=E 0596'+0EE ^c ]

ſ  ſ
ſ
ſ

ſ ſ  .Lſ  ſ   ſ
ſ 
ſ Nſ
 ſ ſ ſſ
ſ 
ſ.
Lſ
ſ$ſſ ſ$ſ ſ.
ſ$ſ
 ſ
;
"E E Nſ

ſ
ſ
ſ1
ſ ¤
ſ$ſ ; 
ſ
ſ ſ
ſ
 ſ
ſ
ſ  ;
ſ.
ſſ
ſ .ſſ
ſ 
 ſ
 ſ  $ ' 5 P 
ſ Kſ ſ .;
-*N D*5=*?"DF@*N :+N "N &:33.*N :+N B:+DN )@/80N m ſſſ 

5ſ ċ
ſ
ſº
ſBN
;  ſ
ſ. ſſ  ſſ
ſ
ſ
6/1*BN "I"LN /BN ?"D-*@N F8B"1"&1*N /8+:@6"D/:8N N 
 ſ œCſ \Cſ ÁCſ [ ſ ſ ſ  1ſ  .ſ

Lſ  ſ
L*DN " 8 N/8BD?F5*8DN I-/'-N ':F1)N ,/H*N L:FN D-"DN D*6M
´™´óſ ſ ſ .ſ ſ ſ  ſ 
ſ   ſ 
ſ ſ a 
ſ $ſ ſ ſ $ſ ſ
=*?"DF?*N-"BN&**8N)*H*1:=*)N&LN18#@ND*'-8/'/"8B N

ſ  
 ſ ſ ſ ſ ſ

ſ *ſ .
ſ$ſ
 ſ

ſ
ſ ſ
ſ
 ſ
:?N L*"@BN 18:@N -"BN 5"/8D"/8*)N "N =@:,@"6N :+N

ſ 
 ſſ
 ſſ
ſ
ſ
ſ 
ſ
ſ  
 5ſ _ſ .  ſ
*K=2:@"D:?LN ?*B*"@'-N N1::0/8,N "DN :1) N=@:&1*6BN
+/8)/8,N 9*IN B:1FD/:9B N"'-N 18#@N /8BD@F5*8DN /BN
 ſ ſ.
¬ſ 
  ÷ſ P 
ſ [ſ9
 ſſ ſ
$N =@:)F'DN :+N D-/BN 5:)*?8N D-/80/8,N N)*B/,8*)N D:N ‹

ſ 
ſ ſ .ſ $ſ
ſ  
 Ý

ſ   ſ 
ſ ſ I$
ſ $ſ ſ Bſ $ſ
6**DN D:)"LBN (:8)/D/:8B N ſ 
ſ €KVLſ ſ 

ſ €[Vſ Nſ  ſ  Mſ .,
jſ  
ſ PLſ ſ
8*N D-*N L@:3;'N ':8D/8F"11LN 5:8/D:@BN F=N D:N m ſ ſ €PVſRſ ſ X1ſ  Þ

ſ   ſ 
ſ ſ 9
ſ $ſ Nſ ſ $ſ
D*6=*@"DF@*N =:/8DBN    I"@8BN
 :+N :H*A-*"D*)N '/AN N .ſ ſ  % ' 5 P 
ſ KLſ  ſ  .ſ.
jſ  
ſKſ

ſ
'F/DBN N=/8=:/8DBN D@:F&1*N B=:DB N-*N 18:@N *IM ¥1ſ
ſ1ſ
 ſſ
ſ
ſ .ſ  ſ1<ſ ſ ſſſ5ſ Kſ  ſ
>:/8D*@N ?*)F'*BN )*IN =:/8DN 5*"BF@*5*8DN D:N "N
$ſ t  <ſ   ſ
ſ . ſ 
ſ ſ ; ſ ſ 1 1ſ  1
ſ $ſ Lſ 
ſ
C/6=1*N *K"'DN =@:'*BB 
N :AD"&1*N"8)NB*1+':8D"/8*)N
+<AN 1"&:@"D:@LN "''FA"'LN "8LI-*@* N -*N -"8)LN
ſ ſ..  ſ 
ſſ ſ ſ ſ $ 
ſ

ſ  
ſ Kſ ;
18:?N L@:':8N D"0*BN "'F@"D*N BGA+"'*N D*6=*@$NN $ſ ſ.
ſ$ſ
ſ  .ſ

ſ a ſ ſ ſ $ſ
ſ 

ſ $ſ
ſ
DF@*BN :+N "8LN 6"D*A/"1N "8LN B-$=* N 
R
ſ  ſ
 R ſ
ſ  M1ſ I
ſ <ſ  
ſ [Aſ
ſ $1ſ
:8H*8/*8'*N /)*8D/+/*BN "11N 11/8:/BN *BD/8,N $&NN Mſ 
ſ $ſ
ſ
ſ 
ſ  
 ſ $ƒſ$ſ
ſ>
ſ Lſ  
ſPLſ
:@"D:@LN =@:)F'DBN %BN BG@*1LN "BN D-*N 8"6*N
18:A N
ſ9
ſ§ ?
ſſ$ſ
ſI ţî ŕőjſ
-*D-*@N L:FN 5*"BF@*N D*6=*A"DF@*N )*IN=:/8DN <@N
 
ſ KLſ
ſ 9
ſ 1
ſ $ſ ſ ſ$ſ
$/@N H*1:'/DLN L:FN ,*DN +"EDN =:B/D/H*N @*$)/8,BN D<=N ]c

ſ  Lſ ſ ſ Lſ  ſ 1ſ
=:@D"&/1/DLN "8)N +/8/D*N "''F@"'L N
9
 ſ 1ſ  M
ſ ſ $ſ
ſ
*8)N D:)"LN +:@N 1/D*@"DG@*N :8N D-*N 18:@N /8BD@GM
7*8DN I-/'-N I/11N B*?H*N L:F@N 8**)BN D-*N 5:)*@8N

 ſ $ſ 
ſ M  5ſ J
ſ ſ ſ ſ
J"L N!@/D*N D:N 11/8:/BN *BD/8,N "&:?"D:@/*BN 8' N

ſ$
ſ
ſ ſſ$ſ
ſ1N*Ž
::5N mm 2 m ""41*N D N -/'",:N m 11/8</B N ſ  Mſ

ſ Ľſ  ſ ;
 jſ  ſ
ſ 
ſ ſ ſ

ſ ſ  ſ  1ſ
ſ
ſ  :


 ſſ ſſ$ ſ

ſſ
ſ ;

 ſ ſ 
ſſ ſ  ſ .
–ſ
]

ſ 
ſ  Mſ
 ſ M
;

<ſ 
ſ 1 ſ   ſ a  ſ şſ
ſ
 .
ſ
5ſ
' '   
' /À,)ċ
$$#'  ' ' ' ""' &%&'  ſ 
&
Óſ
ſ&
ſ
ſ ſž
!$' ' '
Dm .
×ſ 2#ſ ŞÓſ ōÅ#Űſ 2# # åſ Åſ
-å
ċ—ċ Ě ſ ..  ſ ſ ſ ſ N D

ċ

‹6&,(17,),&$0(5,&$1,1&

This content downloaded from

86.60.213.150 on Wed, 29 Sep 2021 14:44:16 UTC
All use subject to https://about.jstor.org/terms
 Thermoplastic man-made fibres can be felted.
American knows how. They can be fabricated
into special parts and products with dimen­
sional stability and accuracy for assembly
or use without further processing. If desired,
fused edge products can be joined, to others
by conventional methods including stitching;
The felt within the edges can have any desired
Figure 7 porosity, within wide limits, since such felts
can be made entirely of manufactured fibres,
or contain mixtures of natural and man-made
counted true by intuition were repeated­ fibres. Th� s these fused-edge felts have great
ly proved false by logic, mathematicians versatility, and are capable of rendering
became more and more skeptical of the many different services. It will pay you to
validity of intuition. The conviction look into what fused-edge felt products and
parts can do for you. Write for information
grew that it was unsafe to accept any
on your company letterhead.
mathematical proposition, much less to
base any mathematical discipline on in­ Some Present Uses
tuitive convictions. Thus a demand arose Vacuum cleaner filters
for the expulsion of intuition from Powdered soap containers
mathematical reasoning and for the com­ Face powder pads
plete formalization of mathematics. That Insulation

is to say, every new mathematical con­ Clothing lining

cept was to be introduced through a Fluid filters

(The field has barely
purely logical definition; every mathe­
been scratched!)
matical proof was to be carried through
by strictly logical means. The pioneers
of this program (to mention only the
most famous) were Augustin Cauchy
(1789-1857), Bernhard Bolzano (1781-
RECORDS
1848), Karl Weierstrass (1815-1897), TWO INDEPENDENT
Georg Cantor (1845-1918) and Julius
VARIABLES FROM
Wilhelm Richard Dedekind (1831-
ANALOG OR DIGITAL
1916). INPUTS
The task of completely formalizing or
logicizing mathematics was arduous and A compact, desk-size unit
difficult; it meant nothing less than a designed for general purpose
root-and-branch reform. Propositions graphic recording from analog
that had been accepted as intuitively evi­ or digital inputs with stand­
dent had to be painstakingly proved. As ard Librascope converters or
special modifications engi­
the prototype of an a priori synthetic
neered to customer require­
judgment based on pure intuition Kant ments. Unique pen travel,
expressly cited the proposition that space fast and dependable., Full
is three-dimensional. But by present-day chart visibility allowing
curve generation to be ob­
served at all times. Write for
detailed catalog information.
Mechanical and electrical
analog computers, digital
computers, input·output
devices and components.

Computers and Controls

0-----'

Figure 9

87

© 1954 SCIENTIFIC AMERICAN, INC

This content downloaded from

86.60.213.150 on Wed, 29 Sep 2021 14:44:16 UTC
All use subject to https://about.jstor.org/terms
 a b c standards even this statement calls for yond that the theory also makes possible
searching logical analysis. First it is neces­ an unusually precise and subtle analysis
sary to define purely logically what is of the structure of curves, about which I
meant by the "dimensionality" of a geo­ should like to comment briefly.
metric figure, and then it must be proved A point on a curve is called an end
logically that the space of ordinary pOint if there are arbitrarily small neigh­
geometry-which is also the space of borhoods surrounding it, each of whose
Newtonian physics-as embraced in this boundaries has only a single point in
definition is in fact three-dimensional. common with the curve [points a and b
This proof was not achieved until 1922, in Figure 16]. A point on the curve that
and then simultaneously by the Vienna is not an end point is called an ordinary
mathematician K. Menger and the Rus­ point if it has arbitrarily small neighbor­
sian mathematician Pavel Uryson (who hoods each of whose boundaries has ex­
Figure 10 later succumbed to a tragic accident at actly two points in common with the
the height of his creative powers) . I wish curve [point c in Figure 16]. A point on
to explain briefly how the dimensionality a curve is called a branch point if the
of a figure is defined. boundary of any of its arbitrarily small
a b c A geometric figure is called a "point neighborhoods has more than two points
set." It is said to be null-dimensional if in common with the curve [paint d in
for each of its points there exists an Figure 16]. Intuition seems to indicate
arbitrarily small neighborhood whose that it is impossible for a curve to be
boundary contains no point of the set. made up of nothing but end points or
For example, every set consisting of a branch points. As far as end points are
finite number of points is null-dimen­ concerned, this intuitive conviction has
sional, but there are also many compli­ been confirmed by logical analysis, but
cated null-dimensional points which as regards branch points it has been
consist of infinitely many points [Figure refuted. The Polish mathematician W.
14]. A point set that is not null-di­ Sierpinski proved in 1915 that there are
mensional is called one-dimensional if curves all of whose points are branch
for each of its points there is an arbi­ points. Let us attempt to visualize how
Figure 11 trarily small neighborhood whose bound­ this comes about.
ary has only a null-dimensional set in Suppose that an equilateral triangle
common with the point set [Figw'e 15]. has been inscribed within another equi­
Every straight line, every figure com­ lateral triangle and the interior of the
a b c posed of a finite number of straight lines, inscribed triangle erased [Figure 17]. In
every circle, every ellipse-in short, all each of the three remaining triangles
geometrical constructs that we ordinarily [the unhatched areas] inscribe an equi­
designate as curves-are one-dimensional lateral triangle and again erase its in­
in this sense. A point set that is neither terior; there are now nine equilateral
null-dimensional nor one-dimensional is triangles together with their sides [Fig­
called two-dimensional if for each of its me 18]. Imagine this process continued
points there is an arbitrarily small neigh­ indefinitely. (Figure 19 shows the fifth
borhood whose boundary has at the most step, where 243 triangles remain.) The
a one-dimensional set in common with points of the original equilateral tri­
the point set. Every plane, every poly­ angle that survive the infinitely numer­
gonic or circular area, every spherical ous erasures can be shown to form a
surface-in short, every geometric con­ curve all of whose points, with the ex­
Figure 12 struct ordinarily classified as a surface­ ception of the vertex points a, band c of
is two-dimensional in this sense. A point the original triangle, are branch points.
set that is neither null-dimensional, one­ From this it is easy to obtain a curve
dimensional nor two-dimensional is with all its points branch points; for in­
a b c called three-dimensional if for each of its stance, by distorting the entire figure so
points there is an arbitrarily small neigh­ that the three vertices of the original
borhood whose boundary has at most a triangle are brought together in a single
two-dimensional set in common with the point.
point set. It can be proved-not at all But enough of examples-let us now
simply, however-that the space of or­ summarize what has been said. Re­
dinary geometry is a three-dimensional peatedly we have found that in geo­
pOint set. metric questions, even in very simple
and elementary ones, intuition is a
his theory provides what we have wholly unreliable guide. And it is of
T been seeking-a fully satisfactory course impossible to adopt this dis­
definition of the concept of a curve. The credited aid as the basis of a mathe­
essential characteristic of a curve turns matical discipline. The way is then open
Figure 13 out to be its one-dimensionality. But be- for other logical constructs in the form

88

© 1954 SCIENTIFIC AMERICAN, INC

This content downloaded from

86.60.213.150 on Wed, 29 Sep 2021 14:44:16 UTC
All use subject to https://about.jstor.org/terms
------- -------

ENGINEERS
the best careers are
made here, too!

Westinghouse Air Arm-world leader in the field of

aviation-electronics-also leads in engineering career
opportunities. Continued growth of operations has
created a need for engineers who would like to establish
themselves in positions with a solid future. In addition
to high income and outstanding benefits, current
openings offer- . IIIiILilliiliili..IiI........illiIiI. Analysis of aircraft tactical

• Job stability • Extensive facilities problems. Evaluation of detection, acquisition,

• Opportunities for advanced study and destruction probabilities of aircraft arma­

ment systems. Study of weapons systems
• Association with competent technical leaders
stability and accuracy requirements.

Flight Test Engineers responsible for laboratory

and flight evaluation of radar directed fire con­ Computer Design Engineers Analysis and design

J
trol, autopilot, and navigation systems. Flight of analog ".nd digital computers for airborne
test these systems in many of nation's newest
fire control systems.
jet intercepters and attack bombers. Gain
knowledge of complete airborne electronics
systems. Hydraulics Engineers Analysis and design of hy.
draulic valves. pistons, actuators and servo­
mechanisms for autopilot and fire control
,11!Jo'I..........iIII�..�IiII�....Iii.. Technical consultation on
systems.
problems pertaining to coherent and non-co­
herent radar systems. Long range planning on
target detection systems. Analysis of radar
system and determination of component para-
Electronic Components Engineers Design of fire
control system components.
accelerometers magnetic amplifiers indicators
rate gyros transformers antennas
Co nsul t i n g vertical gyros modulators receivers
investigation of insulating
materials in potting and encapsulating of If employed at your highest skill in a defense
electronic subassemblies. Development of sub­ industry do not apply.
miniaturization techniques.

SEND TODAY
For your copy of "Gateway

YOU CAN BE SU to an Engineering Future."

�stlnghou:Se
ADDRESS:
R. M. Swisher, Jr.
Employment Supervisor, Dept. A-6
Westinghouse Electric Corporation
109 W. Lombard Street
Baltimore I, Maryland

AIR ARM DIVISION

� . .. �,
" BALTIMORE, MA'RYLAND

89

© 1954 SCIENTIFIC AMERICAN, INC

This content downloaded from

86.60.213.150 on Wed, 29 Sep 2021 14:44:16 UTC
All use subject to https://about.jstor.org/terms
 WEeDED
TERMINAL
LUG
WIRE
o
Figure 14 Figure 15

of spaces differing from the space of ordinary geometry. They appear as soon
ordinary geometry; spaces, for instance, as we reflect upon objects that we had

'�
�ded
in which the so-called Euclidean parallel not thought about before.
postulate is replaced by a contrary postu­ Modern physics now makes it appear
late (non-Euclidean spaces), spaces appropriate to avail ourselves of the logi­
whose dimensionality is greater than cal constructs of multidimensional and
three, non-Archimedean spaces (in non-Euclidean geometries for the order­
which there are lengths that are greater ing of our experience. (Although we have
than any multiple of a given length). as yet no indication that the inclusion of

terminals
What, then, are we to say to the often­ non-Archimedean geometry might prove
heard objection that the multidimen­ useful, this possibility is by no means ex­
sional, non-Euclidean, non-Archimedean cluded.) But, because these advances in
geometries, though consistent as logical physics are very recent, we are not yet
'KOW �te. � oJ! constructs, are useless in arranging our accustomed to the manipulation of these
experience because they do not satisfy logical constructs; hence they are still

CIHIMDT�® intuition? My first comment is that or­

dinary geometry itself is by no means a
considered an affront to intuition.
The same reaction occurred when the
Uli/te-Wt'JUftfi supreme example of the intuitive
process. The fact is that eveTY geometry
theory that the earth is a sphere was ad­
vanced. The hypothesis was widely re­
RESISTORS -three-dimensional as well as multidi­
mensional, Euclidean as well as non­
jected on the grounds that the existence
of the antipodes was contrary to in­
Euclidean, Archimedean as well as non­ tuition; however, we have got used to
Archimedean-is a logical construct. For the conception and today it no longer
Provide a p erfect several centuries, almost up to the occurs to anyone to pronounce it impos­
permanent connection sible because it conflicts with intuition.
present day, ordinary geometry admi­
between resistance rably served the purpose of ordering our
wire and terminal
experience; thus we grew used to operat­ p hysical concepts are also logical con-
Here's the ultimate in resistor ing with it. This explains why we regard structs, and here too we can see clear­
stability-welded terminals that it as intuitive, and every departure from ly how concepts whose applicHtion is
are unaffected by vibration or it contrary to intuition-intuitively im­ familiar to us acquire an intuitive status
high temperatures. This con­ possible. But as we have seen, such "in­ which is denied to those whose applica­
struction eliminates noise or in­ tuitional impossibilities" occur even in tion is unfamiliar. The concept "weight"
stability in highly sensitive cir­
cuits.
c c
Millions of Ohmite welded re­
sistors have already proved their
reliability under the toughest
service conditions.

OHMITE MFG. CO.

3609 Howard St., Skokie. Ill.
(Suburb of Chicago)

Figure 17 Figure 18

90

© 1954 SCIENTIFIC AMERICAN, INC

This content downloaded from

86.60.213.150 on Wed, 29 Sep 2021 14:44:16 UTC
All use subject to https://about.jstor.org/terms
 FigLLre 16

is so much a part of common experience

that almost everyone regards it as in­
tuitive. The concept "moment of inertia,"
however, does not enter into most peo­
ple's activities and is therefore not re­
garded by them as intuitive; yet among
many experimental physicists and engi­
neers, who constantly work with it, mo­
ment of inertia possesses an intuitive
status equal to that generally accorded
the concept of weight. Similarly the con­
cept "potential difference" is intuitive
for the electrical technician, but not for
most people.
If the use of multidimensional and
non-Euclidean geometries for the order­
ing of our experience continues to prove
itself so that we become more and more
accustomed to dealing with these logical
constructs; if they penetrate into the cur­
riculum of the schools; if we, so to speak,
learn them at our mother's knee as we
now learn three-dimensional Euclidean
geometry-then it will no longer occur to If your eye lights up when the challenge is a big one, when
anyone to say that these geometries are the stakes are high, when the only ceiling is your own ability
contrary to intuition. They will be con­
...if you're willing to tackle long, hard work on difficult and
sidered as deserving of intuitive status
as three-dimensional Euclidean geome­ precarious undertakings ... if you're a creative engineer with
try is today. For it is not true, as Kant a gleam in your eye, then this is it! No plush inducements.
urged, that intuition is a pure a priori
Only the opportunity to work with the finest mindpower and
means of knowledge. Rather it is force
of habit rooted in psychological inertia. facilities i,n the whole new world of spaceborne systems-on

a top-priority problem.
c
If it's only a job you want, the woods are full of them. But if
you aloe one of the few who are destined to go far in this indus­
try, you'd be wise to take an engineer's-eye view of the mind­
power and the facilities you'll be working with.
Write to J. M. Hollyday, Director of Employment, Dept. 5-4

.� � THE GLENN L. MARTIN COMPANY

WBALTIMORE
AIR C RAFT MARYLAND •

Figure 19

91

© 1954 SCIENTIFIC AMERICAN, INC

This content downloaded from

86.60.213.150 on Wed, 29 Sep 2021 14:44:16 UTC
All use subject to https://about.jstor.org/terms