124 ANNUAL REPORT 
it is essential to have at least 75 per cent. of calcium carbonate present in 
the mixture, not so much because less lime might not unlock the silicate 
or not attack the free silica, but owing to the fact that the resulting mass. 
after burning becomes hard and stone-like and hence inconvenient for 
reduction. for this reason it was decided to use a standard mixture of 
So per cent. CaCO, and 20 per cent. of the clay be tested, which results. 
in a friable mass easily reduced. ‘This amount of lime also allows for a 
sufficient excess, so that uniformity is obtained as far as the reagent is. 
concerned. | 
In the determination of the insoluble residue of these mixtures suf- 
cient material was taken to correspond to two grams of the clay. In 
this and other series the total lime was determined volumetrically in a 
number of analyses. This was done in order to express the per cent. of 
residue in terms of the actual weight of clay taken, thus making due 
allowance for the lime introduced as reagent, since the mixing of clay 
and lime can never be so intimate as to make it permissible to weigh out 
any amount of the mixture and obtain the actual amount of clay taken by 
calculation, depending upon the per cent. of calcium carbonate originally 
added under the supposition that the mixture is perfect. Later on this 
method was modified in such a way that 2 grams of the dried clay were 
weighed out, transferred to a beaker and made up into slip with water. - 
To this was added the amount of lime corresponding to the required pro- 
portion of I part of clay to 4 of calcium carbonate. The mixture was 
carefully transferred to a porcelain crucible, evaporated to dryness, mixed 
on a sheet of glazed paper with the spatula and returned to the crucible in 
which the sample was burned, being covered with a lid. In this way, of 
course, the determination of limes becomes unnecessary. 
The analysis proper is carried out by taking the sample, treating 
with hydrochloric acid (1:3), allowing it to boil for five. minutes till in 
solution. This liquid is now filtered and washed nearly free from chlorine, 
the residue being washed back into the casserole and treated with the 
alkali solution as indicated under the head of rational clay analysis till no 
more silicic acid is found in filtrate. The residue is then treated with 
hydrochloric acid (1:4) washed free from chlorine, ignited and weighed. 
The main precaution to be observed is to keep the work as hot as 
possible, while yet allowing the silicic acid to settle. The use of long 
stemmed funnels is to be preferred to suction flasks. 
Burning.—For burning the mixture, there was available a dental 
furnace manufactured by the American Gas Furnace Co. which was in- 
tended to be fired with gasoline vapor, but which was modified so as 
to burn ordinary kerosene. For this purpose a 1% inch pipe was intro- 
Guced into the lower cylindrical part of the furnace. The pipe was run 
to a can fastened to the wall and was fed with oil from the reservoir by 
