330 JOURNAL OF THE WASHINGTON ACADEMY OF SCIENCES VOL. 13, No. 14 
CHEMISTRY—The residue from silica in rock-analysis... M. AuRouss- 
EAU. (Communicated by H.S. WAsHINeTON.) 
The usual procedure in rock-analysis, when dealing with the ‘‘main 
portion” is to decompose about one gram of the rock powder by 
fusion with five times its weight of sodium carbonate. The cake is 
brought into chloride solution, the solution evaporated to dryness 
and the silica separated and washed by cold filtration. The silica 
is then ignited in a weighed platinum crucible, weighed, driven off 
by hydrofluoric acid, and the crucible and residue which it invariably 
contains are again weighed after ignition. Hillebrand and Washington 
have both commented on the amount and nature of this residue, and 
Bloor has investigated it in connection with the analysis of clays—a 
somewhat different process from the analysis of rocks.? 
. Both Hillebrand and Washington agree that the residue is of con- 
stant occurrence, that it differs in nature and in quantity with differ- 
ent kinds of rocks, and that it contains chiefly oxides of titanium, iron, 
and phosphorus. The former states that it is quantitatively pre- 
cipitable with ammonia, and that it should contain little if any lime 
or magnesia if the rock has been properly decomposed. This is not in 
agreement with Bloor’s results. His residues were complex in com- 
position, and many of them contained lime and magnesia in notable 
quantity. No statement of the quantitative composition of the 
residues obtained in rock analysis has ever been published. Hille- 
brand states that he has tested residues, after the appearance of 
Bloor’s paper, and is not able to substantiate Bloor’s results. In my 
experience, the behavior of similar rocks is very irregular, even under 
conditions of work subject to little variation. For example, in - 
analyzing a series of basalts from Etna, the residues obtained were 
unlike in quantity and appearance. Taking all precautions to ensure 
complete decomposition, the residue sometimes indicates that this was 
not attained. 
While analyzing the new material (a silicic andesite) which rose 
in the crater of Lassen Peak, Cal., during its period of activity since 
1914, opportunity was afforded to collect and analyze the residue 
from the silica of an andesite. One incomplete and two complete 
1 Received July 12, 1923. 
* Hillebrand, W. F. The analysis of silicate and carbonate rocks. U. 8S. G. &., 
Bull. 700, 1919, pp. 104-105, and 119 (first footnote). 
Washington, H. S. The chemical analysis of rocks, 3rd. Ed., New York, 1919, 
». 146. 
. Bloor, W.R. Journ Am. Chem. Soc., 29, 1907, p. 1603. 
