631.42:549.

651

MEDEDELINGENVANDE L A N D B O U W H O G E S C H O O L
WAGENINGEN.NEDERLAND. 65-9(1965)

ON THE IDENTIFICATION
OF FELDSPARS IN SOILS

D. J.DOEGLAS,J.CH.L.FAVEJEE, D.J.G.NOTA and L. VAN DER PLAS

Department of Geology and Mineralogy,

Agricultural University, Wageningen, The Netherlands

(Received 25-VI-1965)

SUMMARY

The fertility of a soilisto some extent determined by its mineralogical com-

position. Therefore, the analysis of the entire mineralogical composition of a
soil carries rather useful information. Even if the economic situation in a cer-
tain country allows for an unlimited use of all kinds of fertilizers the above
statement remains true, especially for minor elements.
Attention is focussed on a relatively unknown part of the soil matter, viz,
the light mineral fraction with a particle size between 50and 420 \L. Among the
minerals in this fraction feldspars are important. In the light of new develop-
ments in feldspar identification X-ray powder methods and measurements of
indices of refraction are advocated as the most efficient and reliable techniques.
It isexplained that a quantitative mineralogical analysis of this fraction is made
easier if satisfactory concentration techniques can be applied. Moreover, it is
argued that the toilsome counting of particles in index liquids may be partly
substituted by staining and the much easier counting of yellow and blue grains.
A staining technique as well as a method for the separation of feldspars with
heavyliquids,both developed byFavejee, are described. Finallythe systematical
treatment of a sample is discussed. Problems still present themselves in treating
the fraction between 8and 50 \i.

INTRODUCTION

One of the important aspects of a soil isits fertility. Soilfertility is determined

byalargenumber offactors. Amongthesethecontribution ofelementsfor plant
nutrition bythe minerals present in the soil should not be overlooked, though
the economic situation in a number of countries allows for an unlimited use of
Meded.Landbouwhogeschool Wageningen 65-9(1965) 1
all kinds of fertilizers. This very function of minerals in soils completely war-
rants an attempt at analysing the entire mineralogical composition of soil
samples.
For a systematic treatment the soil matter may be divided into a part con-
sisting of minerals exclusively and a part holding the organic matter. The
mineral fraction of the soilis composed of:
a. a fraction consisting of light minerals, viz, minerals with a sp. gr. < 2.89,
b. a fraction consisting of heavy minerals, viz, minerals with a sp. gr. > 2.89,
c. a fraction consisting of a.o., the so-called clay minerals, the fraction smaller
than 0.5 micron.
Becausethefractions b and chavebeen studied intensively over a long period
of time and by a large number of soil scientists, one would expect their contri-
bution to the soil fertility to be well known at present. However, the clay frac-
tionand the light fraction may well be a much larger source of minor elements
suchascopper,manganese,magnesium and borium than thenegligible amounts
of such elements present in the heavy fraction.
In comparison with these two the light fraction is still a rather unknown part
ofthe soilmatter, notwithstanding theefforts ofa largegroup ofworkersin this
field of investigation. The difficulty of quickly distinguishing between quartz
and feldspar, and between different feldspars is one of the reasons for this
fragmentary knowledge. We made a rather intensive study of the methods for
the determination of these minerals. By a proper selection and by modifying
existing methods a set of routine techniques has been developed. Such routine
methods must rapidly give adequate information on the mineralogical compo-
sition of the light fraction of a soil sample. In the following we will describe
the considerations pertinent to this selection and modification.

THE COMPOSITION OF THE LIGHT FRACTION

Any textbook of mineralogy may inform us about the minerals that may
theoretically be found in the light sand fraction of a soil sample. Analyses of a
TABLE 1. Specificgravitiesofanumberofmineralstobeexpectedinthelightfraction ofsome
soils or sediments

Opal 1.95-2.10 Beryl 2.65-2.85

Sodalite 2.2 -2.4 Calcite 2.71-2.72
Gypsum 2.2 -2.4 Dolomite 2.87
Serpentinite 2.5 -2.65 Prehnite 2.80-2.95
Alkali feldspar 2.56-2.62
Cordierite 2.57-2.66 Chlorite 2.6 -3.0
Chalcedony 2.6 -2.65 Stilpnomelane 2.7 -3.0
Scapolite 2.61-2,70 Muscovite 2.76-3.00
Plagioclase 2.62-2.76 Biotite 2.8 -3.4
Quartz* 2.63-2.67

* The specific gravity of quartz mayvary considerably due to small inclusions of gas, liquid
and/or heavy minerals.

2 Meded.Landbouwhogeschool Wageningen 65-9(1965)

large number of samples from sediments and soils make it clear that q u a r t z ,
f e l d s p a r , c a r b o n a t e s and m i c a are frequently important constituents of
this fraction. Other minerals, cf., table 1, may occasionally be found.
The determination of the CaCO s content is comparatively easy. The analysis
ofthe quantity ofmica may be done by sorting methods based on the difference
of morphology and by magnetic separation, in the same way as described for
the preparation of mica concentrates for age determination by the Rb-Sr
method, cf., EMILY JÄGER (1962). The identification of feldspars in the light
fraction of soils remains to be treated. The identification of quartz goes along
with that of feldspars.

FELDSPARS

Thelatest developments in the study of the different phases that belongto the
feldspar family necessitate a revision of the conception of older textbooks. It
would lead too far to discuss the details of the present situation, the more so
because a book on detrital feldspar identification for sedimentologists and soil
scientists is in press (VANDER PLAS, 1965). For those unfamiliar with the new
ideas the following rather simplified feldspars scheme may be of interest, cf.,
table 2.

TABLE 2. An explicative scheme illustrating the more simple relation between the various
feldspar phases known tooccur frequently inrocks.

Chemicallyfeldspars are crystalline phases offormula AB 4 0 8 ;

A, alarge low valent cation e.g., K,Na,Ca.
B, includes both Al andSi.
Chemically feldspars may thus beclassified:
K AlSis08, abbreviated Or(from orthoclase) : Potassium feldspar
(KNa)AlSi808, : Alkalifeldspar
Na AlSi308, abbreviated Ab(from albite) : Albite
(NaCa)AlSi308, : Plagjoclase feldspar
CaAlaSiaOg ,abbreviated An(from anorthite) : Anorthite
Physicallyfeldspars form a system ofvarious phases with a continuous structural variation
between high- andlow-temperature modifications. Thelowtemperature binary systemsof
alkalifeldspars and ofplagioclasesdonotform acontinuous seriesofmixedcrystals between
the endmembers; unmixing maybe seen in some of the chemically intermediate phases.
Perthites and peristerites arethe result ofsuch unmixing processesinthe alkali feldsparsand
in the plagioclases respectively.
Alkali feldspars:
High temperature phases, monoclinic : sanidine*
Low temperature phases, triclinic : microcline
Perthites, lamellar aggregates formed ofalkali feldspar andsodium-rich feldspar.
Plagioclase feldspars:
High temperature phases, triclinic continuous series ofmixed crystals
Low temperature phases, triclinic discontinuous series of mixed crystals, with various
structures.

* Suchnamesasorthoclase,anorthoclase,adularia etc. shouldnowbeavoidedinsoilscience.

Meded.Landbouwhogeschool Wageningen 65-9(1965) 3

 FELDSPARS IN SOILS

It must be stressed that the feldspar assemblage of a soil is fundamentally

different from that of an igneous or metamorphic rock. Methods for the study
of feldspars, however, have been developed to a large extent by students of
these igneous and metamorphic rocks. Therefore it is conceivable that these
methods need some modifications.
Themaindifference betweenthefeldspar content ofigneousand metamorphic
rocks on one hand and of soils developed on recent or old sedimentary deposits
on the other hand is due to the fact that the mineral assemblages of the first
group of rocks are governed bythe laws ofthermodynamics, whereas the mine-
ral assemblages of soils are partly a function of other factors such as:region of
provenance of the parent material; the various stabilities of minerals under
weathering conditions; reworking of sediments and mixing of sediments from
different sources. In igneous and metamorphic rocks, a limited number of
minerals are found next to each other. The investigator of these rocks assumes
from experience that the results of a few determinations of feldspar fragments
may be generalized for the whole complex. In soils developed on sedimentary
rocksthe feldspars present may have derived from an u n k n o w n n u m b e r of
u n k n o w n r o c k s . Acid and calcic plagioclases as well as high- and low-tem-
perature phases of alkali feldspars and plagioclases may be found next to each
other in one sample.
A soil sample is, furthermore, a collection of large and small mineral parti-
cles. Feldspars are to be expected both in the small and in the larger size frac-
tions. The methods for the study of large feldspar particles differ considerably
from those for studying small feldspar particles. Still determinations have to
be made in every size fraction, because the feldspar assemblages may differ
from size fraction to size fraction. Moreover, the information of different
methods is not always comparable.
Finally the amount of feldspar particles in a soil may well be rather low.
For that reason it is difficult to make an efficient use of the available determi-
nation methods, especially if one wants quantitative information. Therefore,
we make RETGERS' (1891) principle about the necessity to concentrate certain
minerals before they are studied, the keystone of our procedures.

CONCENTRATION METHODS
Different methods have been used in order to identify thefeldsparsinthelight
fraction of soils. Determination of feldspars can be a rather time-consuming
occupation. Therefore, the investigator does not like to waste much of his time
inspecting large numbers of other particles. Consequently, a first step in setting
up a routine procedure for feldspar identification is the development of an
adequate concentration method.
In mineral industry flotation techniques and electrostatical separation are
successfully used for the concentration of feldspars, cf., KIRCHBERG (1955). For
large samples flotation is also used on a laboratory scale in many ore-dressing
4 Meded.Landbouwhogeschool Wageningen 65-9 (1965)
laboratories. Preliminary investigations showed that soil scientists may expect
good results from flotation of even small samples in the near future.
A well-known method for the concentration of feldspars is based on the
specific gravity of these minerals. In the following sections we will deal with the
specific gravity concentration of feldspars in detail. The other methods des-
cribed before will not be discussed any further.

SPECIFIC GRAVITY SEPARATION OF FELDSPARS

Specific gravity concentration of feldspars in different fractions has already

been usedby RETGERS(1891),although theliquidsheused at that timehave now
become obsolete. In specific gravity separation with liquids two aspects must be
considered. The first deals with the shape and the size of separation vessels. The
second aspect concerns the properties of the liquids and the particles therein.

Separating vessels.
After alargenumber ofexperimentswithdifferent types ofvesselsand funnels
Favejee developed a special type of funnel with steep walls for the specific
gravity separation of feldspar particles-with a minimum size of 100 microns,
cf., fig. 1. Particles smaller than 100 microns are efficiently separated into
different specific gravity fractions by making use of a centrifuge. Centrifuge
separations, however, are delicate to perform if the specific gravity difference
betweentheparticlesand theliquid is rather small. Numerous suggestions as to
the shape of the tubes and the specialtools to be used for securing the separates

-H 35mm h-

FIG. 1. Favejee funnel. The measure gives an impression of the size.

The drawing is true to scale.
Meded. Landbouwhogeschool Wageningen 65-9 (1965)
have been made. It will be found that tubes with stoppers, pipettes and plastic
bags are difficult to handle in feldspar separation, because remixing occurs
rather easily, not to mention particles that remain in suspension. Thetube de-
scribed by VAN DER VEEN (1963, p. 44, Fig. 3-2) and presumably designed by
Taylor is rather easy to handle.
Heavy liquids
The practical limitations of specific gravity separation are for a large part
due to the various properties of the liquid. These properties are systematically
listed in the following enumeration:
a. The viscosity of the liquid causes small particles to settle rather slowly.
b. The small differences between the specific gravity of the different feldspars
require an accurate determination of the specific gravity of the liquid.
c. Some of the available feldspar particles may have exactly or approximately
the same specific gravity as the liquid; they consequently stay in suspension.
d. The occurrence of specific lighter or specific heavier inclusions in feldspars
such as albites influences the specific gravity of the entire particle.
e. Variations of the temperature of the heavy liquids cause variations in the
specific gravity.
f. Liquids obtained bymixing may change their propertiesinthe course of time
due to the different vapour pressures of the components of the mixture.
From the foregoing enumeration it follows that a useful liquid must have a
low viscosity. If it is a mixture of liquids the components must have approxi-
mately the same vapour pressure. Favejee advocates a mixture of bromoform
and decaline. This mixture has good wetting properties and thevapour pressure
of the components is almost the same. Our experiments show this liquid to be
extremely stable over periods of more than two years.
Mixtures of liquidsfor use in routine procedures.
The feldspars of the light fraction must be divided into a few subfractions.
Every subfraction must hold a group of feldspars. The group limits must be set
in such a way that they meet the requirements of the subsequent mineralogical
analysis. The limits also depend on the specific gravity differences that can
still be used efficiently in a separation with heavy liquids. We found that the
difference between the specific gravity of the liquids must be at least 0.04.
The light fraction is generally rich in quartz. RETGERS (1891) already ob-
served that an abundance of quartz hinders the identification of other minerals.
Therefore, the specific gravity fraction holding the bulk of quartz particles
must be separated from the rest of the sample. The apparent specific gravity
of quartz particles ranges between 2.63 and 2.67. Unfortunately oligoclases
are also found within this range.
It is only too obvious that we aim in the first place at the production of an
alkali feldspar fraction and a plagioclase fraction. Therefore, one of the liquids
must have a specific gravity of 2.59, the maximum specific gravity observed in
alkali feldspars.
6 Meded. Landbouwhogeschool Wageningen 65-9 (1965)
 On thebasis ofthese considerations Favejee suggested thedensitieslistedin
table 3. NOTAand BAKKER(1960)proved this suggestion to be ofpractical value.
In thecourse ofour investigations since 1960,theusefulness ofthese limitshas
been experienced in numerous routine analysis of light fractions of samples
from sands and soils. Finally onemust be aware of the possible presence of
other minerals inthefeldspar concentrates obtained inthewayjust described.
These minerals may befound inthelist oftable 1.

TABLE 3. Density limitsfortheseparation offeldspar fractions with heavyliquids

Sp.gr.,2.59; the lighter fraction holds alkali feldspars exclusively

2.59- 2.63 ;thefraction holding albite andsome quartz
2.63- 2.67;the fraction holding quartz andacid plagioclases
2.67- 2.89;the fraction holding calcic plagioclases

IDENTIFICATION METHODS

In this section wewill deal with three types of identification methods.The

first one isbased ona selective staining offeldspars asa whole anda selective
staining ofalkali feldspars. This method enables anevaluation ofthetotal feld-
spar content, of the alkali feldspar content and consequently of the quartz
content and theplagioclase content bysubstraction. Moreover staining methods
enable theanalyst toprepare small concentrates by handpicking, provided the
grain size isfairly large.
The second method isbased onoptical properties ingeneral andonindexof
refraction andaxial angle in particular. A third deals with theX-ray powder
method.
In petrography staining techniques are used on a large scale in order to
facilitateidentification. Staining offeldspars inthin sections isanoldtechnique
and especially useful if these minerals form a small part of the sample only.
Favejee developedaroutinemethod for staining feldspar particles with a mini-
mum size of 50 y.. The method is based ontheformulas of GABRIEL and Cox
(1929) and of REEDER and MACALLISTER (1957). The coatings produced on the
mineral surface by hydrofluoric acid vapour are stained with different chemi-
cals. Alkali feldspars arestained yellow with sodium cobaltinitrite forming the
yellow potassium cobaltinitrite with the potassium ionsinthe etch coating.The
aluminium ion present in the coatings reacts with hemateine forming a pur-
plish blue compound. Ifboth staining methods aresimultaneously appliedthe
cobaltinitrite staining method must precede the hemateine staining. It is per-
haps not superfluous to saythat the etch coatings on quartz particles remain
unstained, hence the identification of quartz particles from the absence of a
coloured coating in samples stained with hemateine.
We usea small platinum dish, with a flat bottom (fig.2 G).The particles
should just about cover the bottom of the dish in a layer of single particles.
Only then canthe etching vapour and the wetting with solutions afterwards
Meded. Landbouwhogeschool Wageningen 65-9(1965) 7
 —J30mm —

ü
T^
130 mm

100mm-
o
O o
0 0
G o O

H,

FIG. 2. Vesselfor etchingfeldspar particles with HF-vapour. A; the plastic crucible holding
HF-liquid. B; the support, made of polyethylene, for the platinum dish G. C; a tin
mantel supporting the plastic vessel with openings F., for the escaping steam of the
waterbath. D; thelid of theetching vessel. E; the stopper made op plasticclosing the
hole in the lid D. H; boiling stones. The measures are approximate and give an im-
pression ofthesizeofthedifferent parts.

easilyreachthesurface ofeachparticle. Etchingisdoneina plasticvesselwith

a lid, fig. 2D. The vessel containing a small plastic crucible with the HF-
liquid,fig.2A, issurroundedwitha water mantel. Thewater iskept at boiling
temperature. Thelid ofthevesselhasa smallhole,againcoveredwithaplastic
stopper,fig.2E.Asupport for thesmallplatinum dish,fig.2G, is attached to
the stopper aspictured infig.2B. The platinum dish isplaced on the support
and lowered in the etching vessel. The particles of feldspar are kept in the
vessel at a temperature of about 90°C for exactly one minute. The stopper
closesthe vessel during this time. The coating produced in this way is slightly
hygroscopic and only loosely connected with the particle surface. Therefore,
Favejee heatsthe sampleina smallfurnace at 400°Cfor about 5minutes.The
coatings are now more strongly fixed on the surface. In this way staining is
8 Meded. Landbouwhogeschool Wageningen 65-9(1965)
more easily performed, the quality of the stain isbetter and the particles do not
tend to agglomerate.Staining is done in much the same way as described in the
paper of REEDER and MACALLISTER (1957). For the interested reader the for-
mulas follow here in a concise form.
Staining of feldspars
R e a g e n t s : A., a solution of 50 mg of hemateine in 100 ml of 95% ethanol.
B., a buffer solution consisting of 20 g sodium acetate in 100ml of distilled
water, to which 6ml of glacial acetic acid is added afterwards. Dilute the whole
to 200 ml. The solution is approximately 0.5 N and is buffered at pH 4.8.
P r o c e d u r e : The sample after etching and heating in the furnace is covered
with 10drops of the hemateine solution A.,and 5drops of the buffer solution
B. The wholeis immediately mixed by swirling the platinum dish for about one
minute. Five minutes after the addition of the liquids, the sample is washed
with 95% ethanol. The liquid is siphoned off and the sample is twice washed
with acetone. The feldspars now show a purplish blue stain.
Staining of alkali feldspars
R e a g e n t s : A solution of 1gof sodium cobaltinitrite in 4ml of distilled water.
P r o c e d u r e : The sample in the platinum dish, after etching and heating in the
furnace is covered with the cobaltinitrite solution for one minute. After-
wards the sample is washed with distilled water. The excess of water can be
siphoned off. The alkali feldspars show a yellow stain.
N.B., If one wants to stain the samples with cobaltinitrite and hemateine
simultaneously, thecobaltinitrite staininghas to precede thehemateine staining.
On the other hand, staining can also be done on samples that have previously
been stained and cleaned. For the removal of the cobaltinitrite stain as well as
the hemateine stain we treat the particles for some time in a small porcelain
crucible in one ml of a solution of HCl 2N, to which a few drops of H 2 0 2 30%
are added. The crucible is placed on a waterbath at boiling temperature. In this
way the stain is removed together with the etch residue. The sample can now be
etched and stained anew.
Identification with optical methods
Numerous papers have been written on the optical identification of feldspars.
In our opinion the determination of theindex of refraction givesthe best results
for the identification of detrital feldspars. Alkali feldspars can be identified by
their low index of refraction, plagioclases can even be separated into a few
groups according to the chemical composition. The choice of the proper liquids
for such an identification procedure is dealt with by NOTA and BAKKER (1960).
Measurements of the orientation of the indicatrix though more accurate, are
time-consumingandrequire additional apparatus in most cases. These methods
are not fit for routine procedures. Moreover, the last mentioned technique can
only successfully be applied with thin sections and not with grain slides. Finally
routine work on the basis of optical properties, e.g., on index of refraction
measurements, is dependent on particle size; if the particles are too small, an

Meded. Landbouwhogeschool Wageningen 65-9(1965) 9

efficient identification is out of the question. A distinction between alkali feld-
spars and plagioclases can still be made by phase contrast microscopy, as
witnessed by the results of SCHÖNER (1960). For routine identifications a mini-
mum size of about 35\iseemsa necessity. If werealize that X-ray powder work
on feldspars is restricted by a maximum particle size of about 8[x,it is clear at
once that the main problem of a quantitative feldspar analysis of a soil sample
is met with in the size fraction between approximately 8and 35 JA.Some more
willbe said about this aspect in the section on X-ray powder analyses.
Thestructuralpropertiesoffeldsparsmaybeofinterestfortherateof weather-
ing of these minerals. It is known, that high temperature feldspars such as
sanidine weather more rapidly than low-temperature feldspars such as micro-
cline. Since MARFUNIN (1961) brought evidence of theassumption that mono-
clinic alkali feldspars are generally characterized by axial angles smaller than
44°, it ispossible to distinguish between high-temperature and low-temperature
alkali feldspars. From now on alkali feldspars with an axial angle larger than
44°have to betermed microcline, evenifmicroscopical inspection does not give
further evidence for the presence of a triclinic phase. The presence of cross-
hatched twinning or oblique extinction on (001) is not the only diagnostic
property for the identification of triclinic alkali feldspars.

X-ray powder methods

For the study of feldspars in soil samples X-ray methods may have advan-
tages. In some cases the available amount of sample can be rather small, viz, a
specific gravity separate or a small number of grains isolated by hand picking
from a stained sample. It is experienced that the samples are in most cases too
small for the standard sample holders of most diffractometers. For that reason
werely on cameras. For extremely small samples,e.g., 5grains,a small Debeye-
Scherrer camera (FAVEJEE, 1939) is used with a sample prepared according to
HIEMSTRA'S method (1959). If the samples are larger the quadruple Nonius
Guinier-de Wolff camera, also advocated by GOLDSMITH and LAVES (1954,
1961) for feldspar work, is used.
Powder patterns prepared with this camera permit an easy distinction
between sanidines, microclines and perthites. The identification of the various
plagioclases is difficult and from the results of PETERSON and GOLDBERG (1962)
with fine-grained marine sediments, it can be concluded that the identification
methods developed by SMITH (1956) and SMITH and GAY (1958) give the best
performance.
Experiments with feldspar samples taught us, that the maximum particle
size giving reproducible results in X-ray powder patterns is about 8 [i. The
quantification of feldspar particles in silt size samples with a microscope is
efficiently done if the particle size is over 35 [i. The problem remains how to
determine the amount of total feldspar as well as the amount of the various
types of feldspars in samples with a particle size between 8and 35 \i.
An efficient method to treat this fraction has not yet been found. The mine-
rals in this fraction can be powdered and quantitatively analysed with X-ray
10 Meded.Landbouwhogeschool Wageningen 65-9 (1965)
methods. The main drawback of this procedure is the fact that, because of dif-
ferences inthoughness and theperfect cleavagespresentinfeldspars, the powder
thus prepared holds more fine feldspar material than fine quartz. To avoid this
disadvantage one may powder the sample by hand in an agath mortar under
acetone using small amounts of sample at a time.

ASPECTS OF A TENTATIVE ANALYTIC SCHEME

It is self-evident that the discussion of the entire mineralogical analysis of a
soil sample is out of the question. Werestrict ourselves to those parts for which
we developed a new combination of methods.
To begin with the pretreatment of the sample, the organic matter is removed
by a 10%H 2 0 2 treatment over a water-bath. In order to remove the carbonates
the sample is treated with cold 0.2 N HCl. The iron compounds covering the
particles are removed with sodium dithionite. After these treatments the sample
iswetsieved over a 50micron sieveand over a 420\isieve,the parts are weighed
and retained.

Thefraction 50-420 microns

The analytical results secured from this fraction are used as a source of pre-
liminary information. The heavy minerals are separated with bromoform,
weighed and counted. The light fraction iscarefully collected and weighed. This
fraction is afterwards stained with hemateine as described before. The stained
sample, usually about 10mg.,isinspected with a stereomicroscopeand incident
light in order to estimate the total feldspar content and in order to get an im-
pression about the distribution of feldspars over the particle sizes. The magni-
fication usually does not exceed 40x.The same sample after cleaning in the way
described is stained yellow with cobaltinitrite for an estimate of the alkali feld-
spar content. Combining the results lead to an insight in the feldspar content of
the sample. If necessary these samples can be counted.

Thefractions 420-210 (iand 210-105 \i.

The fractions 420-210 \L and 210-105\iare treated similarly. Firstly the heavy
minerals are collected, weighed and counted. The light minerals are used for
evaluating the amount of feldspar. Each size fraction is treated with heavy
liquids in Favejee funnels as described before. Afterwards the specific gravity
fractions are used as follows: (cf. NOTA and BAKKER, 1960, Fig. 1).

The alkalifeldspar fraction, specific gravity < 2.59

If the specific gravity separation has been carefully performed, this fraction
holds alkali feldspars and perthites exclusively. The sample is first weighed and
then stainedyellowwith cobaltinitrite and counted. Itis experienced that apart
from feldspars the fraction may hold flint fragments catching a faint yellow
stain. Inspection with incident light enables the analyst to distinguish between
feldspars and flint fragments. In questionable cases it is advisable to separate a
few suspicious fragments, clean them with HCl and inspect them with a petro-
Meded. Landbouwhogeschool Wageningen 65-9(1965) 11
graphiemicroscope. If the various properties of stained particles are frequently
inspected in this way, the analyst will soon get experienced.
Partly stained particles occur frequently. The proper nature of these grains
may be ascertained by means of microscopic inspection after cleaning. Another
method is to isolate and clean a few fragments, and stain them purplish blue
with hemateine. If albite-alkali feldspar aggregates (perthites) are present, the
fragments are coloured throughout; quartz-alkali feldspar aggregates still show
partly coloured particles.
The nature of the feldspars can further be determined by optical or X-ray
analysis. One has to be aware of the fact that the optical data and the X-ray
powder patterns of alkali feldspars need not give similar results. Under the
microscope a fragment may well appear homogeneous, fully transparent,
untwinned and suggesting the presence of a sanidine, whereas X-ray analysis of
the same fragment betrays its perthitic nature, and makes it evident that an
aggregate of microcline and albite is present.
The albite-quartzfraction, specificgravity 2.59-2.63
This fraction normally holds quartz fragments also and may still hold alkali
feldspar-quartz aggregates and flint. Therefore, the weighed fraction is stained
first yellow in order to spot the alkali feldspars, afterwards the cleaned sample
is stained with hemateine for an estimate of the total feldspar content. Here too
problems are solved with isolated and cleaned grains as described in the former
section.
The quartz plagioclasefraction, specificgravity 2.63-2.67
Soilsamplesgenerallyarerichin quartz. This fraction holdinga large quantity
of quartz has a rather subordinate feldspar content. The weighed fraction hol-
ding plagioclases exclusively, is stained blue and counted. A number of stained
fragments is isolated, cleaned and inspected in index of refraction liquids. This
is necessary because plagioclases, especially those of metamorphic rocks quite
often contain numerous inclusions and may be found in the wrong fraction.
The calcicplagioclasefraction, specificgravity 2.67-2.89
Dutch soils are virtually devoid of calcic plagioclases. In literature, however,
reports are given of soils rich in andesine and even labradorite. The samples
may be stained blue and counted. In one of our experiments on a river sand
from a Swiss tributary of the Rhine, viz the Valserrhein, a large fraction with
specific gravity over 2.67 was found. From a petrographical description of the
region we knew that only albite-bearing metamorphic rocks occur. Optical
inspectionofthefeldspar fragments showed albites with numerous inclusions of
epidote and hornblende. This example may caution not to rely on staining
methods only; a control with other methods is necessary.
Thefraction 105-50 microns
Our experiences with centrifuge methods for specific gravity separation are
rather limited. Some of these methods, however, turned out to be inefficient for
12 Meded. Landbouwhogeschool Wageningen 65-9 (1965)
a routine analysis. At present we treat this entire light fraction with staining
methods only; adequate optical or X-ray control gives these results a higher
reliability.

Thefraction 50-8 microns

This fraction is our main problem and up till now efficient methods did not
result from our experiments. Therefore, we rely in general on time-consuming
optical determinations in the appropriate index of refraction liquids.

CONCLUSION

From the methods reported and from the information about our experiments
it may by clear, that specific gravity separation and staining reactions, though
not used exclusively as means of identification, provide rather quick results
about the constituents of the light fraction of a soil sample. The concentration
technique provides us with fractions in which only a limited number of feldspar
types occurs. This limited number isexperienced as a real contribution towards
the efficiency of thispart of the mineralogical analysis,although we still need to
check our results with optical methods and X-ray photographs.

REFERENCES
FAVEJEE, J. C H . L., 1939. 'Zur Methodik der röntgenographischen Bodenforschung.' Z. Krist.,
A, 100, p. 425-436.
GABRIEL, A. and Cox, P., 1929. 'A staining method for the quantitative determination of
certain rock minerals'. Amer. Miner., 14, p. 290-292.
GOLDSMITH, J. R., and LAVES, F. 1954. 'Potassium feldspars structurally intermediate between
microcline and sanidine'. Geochim. et Cosmochim. Acta, 6, p. 100-118.
GOLDSMITH, J. R.,and LAVES, F. 1961. 'the sodium content of microcline and the microcline-
albite series'. Curs, y Conf., Spain, 8, p. 81-96.
HIEMSTRA, S. A., 1959. 'An easy method to obtain X-ray diffraction patterns of small amounts
of material'. Amer. Miner., 41,p. 519-521.
JÄGER, E., 1962. 'Rb-Sr age determinations on micas and total rocks from the Alps'. J. Geoph.
Research, 67, p. 5293-5306.
KIRCHBERG, H., 1955. 'Feldspataufbereitung'. Montan Rdsch., 3, 290-293.
MARFUNIN, A., 1961. 'The relation between structure and optical orientation in potash-soda
feldspars'. Curs, y Conf., Spain, 8, 111-124.
MARSHALL, C. E., and JEFFRIES, C. D., 1945. 'Mineralogical methods in soil research: Part i.
The correlation of soil types and parent materials with supplementary information on
weathering processes'. Proc. Soil Sei. Soc. Amer., 10, 397-405.
NOTA, D . J. G., and BAKKER, A. M. G., 1960. 'Identification of soil minerals using optical
characteristics and specific gravity separation'. Med. Landbouwhogeschool, Wageningen,
Nederland, 60, (11), 1-11.
PETERSON, M. N . A., and GOLDBERG, E. D., 1962. 'Feldspar distribution in South Pacific
pelagic sediments'. J. Geoph. Research, 67, 3477-3492.
PLAS, L. VAN DER, 1965. 'The identification of detrital feldspars'. Elseviers Publish. Cy.,
Amsterdam, London, New York.
REEDER, S. W., and MCALLISTER, A. L., 1957. 'A staining method for the quantitative deter-
mination of feldspars in rocks and sands from soils'. Canadian J. Soil Sei., 37, 57-59.
RETGERS, J. W., 1891. 'De samenstelling van het duinzand van Nederland'. Nat. Verh. Kon.
Akad. Wet. Amsterdam, Ser. D . i, deel 29, 48 p.

Meded. Landbouwhogeschool Wageningen 65-9 (1965) 13

SCHÖNER,H., 1960. Über die Verteilung und Neubildung der nicht-karbonatischen Mineral-
komponenten der Oberkreide aus der Umgebung von Hannover'. Beitr. Mineral. Petrogr.,
7, p. 76-103.
SMITH, J. V., 1956. 'The powder patterns and lattice parameters of plagioclase feldspars. I.
The soda-rich plagioclases'. Miner. Mag., 31,47-68.
SMITH, J. V. and GAY, P., 1958. 'The powder patterns and lattice parameters of plagioclase
feldspars. II. Miner. Mag., 31, 744-762.
VEEN, A. H. VAN DER, 1963. 'A study of pyrochlore'. Verh. Kon. Nederl. Geol. Mijnbouwk.
Genootschap, Geol. Ser. 22, 189 p.

14 Meded. Landbouwhogeschool Wageningen 65-9 (1965)