6 
\V. F. Cole. 
I’luloubtodiy tlu'ii tiio iargo-sized particles iisod in the o])tical oxaniina- 
tioii wort* not original unbroken partick'^, but on the contrary, were aggregates 
of soil colloid particles (th(^ ]iarticleK being of 2jLt an<l less in ed'ective diameter) 
\vhi('b ha<l formed al’bn’ tb(' dis])ersion of tke mat<'rial. 
OBSFdtVA'flOXvS 
ON 
PHO('KSS OF A(UJKF(JATION OF SOIL 
COLLOIDS. 
A s('ries of obst'rvations \ver(' made with a vi(n\- to determining tlu' par- 
tic'ulai' stage at which tlu' soil colloi<l suspcaision tfnidcd to aggrt^gate. In this 
study (which ^\'as carritnl out on tiu' r('-disp('rs('d material of A 135!) e(.)lloid 
])articl(^s in susptaision w(‘r(' allowtul to s(‘ttle on a fiat surface (under x^arious 
conditions) and the collectt'd niatei-ial was kejit constantly uiuku' examijiation, 
as th(‘ wiutiM' cont(‘nt ^\■as slo^\'ly r(‘mov<'d by ('\aj)oration. It was noted, 
firstly, that th<> [)arti('k's stdl led in siudi a maniuu* tiuit all the material colloetiMl 
remained dark i)c‘tv\'e(Mi crossinl nicols no mattcu' wh(dh(M‘ <>vaporation was slow 
or fast, whetlua* tlocculatcxl or unflocculati'd material was u.sed. Secondly, 
the detinition of tlu‘ interference figure improved as th(‘ material lost water. 
'Llu^ intiatcrc-nci^ figure biM'.auK* only iletlnitely recognizable when the sample 
first showinl tlu' production of contraction cracks. This is in (3onfoimity with 
the ol)S('r\ at ions of other worki'rs (iirindously noted) that indix idual partick's 
(Existing in a soil colloid sus])(‘nsion t(>nd to orient thon.\s(d\'es aftei' drying into 
aggri?gat(‘s whiidi possess uniform optical pi'operties. That thi' whole surface 
was s(*t in an optically uniform manni'r was seen from tlu' fact that the inter- 
feriavce figun' produciKl was th(^ sami' fo?’ all the material collected. This is 
com})let(dy in accord with (Irim's obs('r\-at ions (7) on the pro<hu*tion of ordered 
aggregates after the drying out of clay particles which had beim allowed to 
settle on a sli(k‘ inmu'rsi'd in a clay suspiaision. 
'The inl('i’ri*rence figure (whu-h was |.uirtituilarly good for that material 
which luul lost the largest- amount of watf'i’) had th(‘ saint' charai'tt'ristics as 
are summarist'd in Table 2 for both oi'iginal sampk's of A I34H and A 1351. 
These propertit^s indicate that tlie aggn^gates art' supta’imjioset-l basal plates 
(with jS y), normal to which is tlu* acute bisectrix (t/). In xit'v of the 
X-ray conclusions that tht' soil colloid under (‘xamination (A 1351) contained 
two t'lay minerals, mica aiul kaolinilc, llu'st' oliservations ])rovide addit-ional 
evident't^ that tw*t> cla_\' miiu'i'als bf'kmging to diPiV'rt'nt groups may bt' clostdy 
intergi'ow'u into aggregates which possess unifoi'inity of crystallographic 
orientation. 
SUMMARY. 
An account is given of the results of a niicrosco]'ic examination of soil 
colloid materials from M estern Australia and Tasmania. These results are 
consistent with the luineralogical composition as deduced from X-ray analysis. 
The methods employed and the obsoiw'ations made are in conformity with 
those of other workers in this field. 
At'KXOWLFDrLMENTS. 
These investigations were carried out in the Physics Department of the 
University of Western Australia during the tenure of a Hackett Researcli 
Studentship and a Uommonwca-lth I'csearch grant. The autlior wishes to 
express his thanks to the 1 nive-rsity of Western Australia for the former and 
to th(' Council for Siuontific and rndustriid Reseavi’ch for the latter. 
The author s thanks are also due to Dr. R. T. Pricier, of the Department 
of Geology, and to Mr. J. Shearer, of the Department of Physics, for helpful 
criticism, and to Mr. Burvill, of flie State Department of Agriculture, for the 
redispersion of sample A 1351. 
