332 JOURNAL OF THE WASHINGTON ACADEMY OF SCIENCES VOL. 13, No. 14 
The complexity of the composition of the residue raises the question 
of its recovery. Fortunately this offers no difficulty and takes place 
during the subsequent course of analysis. The precipitate thrown 
down by ammonia in the filtrate from silica is ignited and weighed 
in the same crucible as the residue, when fusion with potassium 
pyrosulphate renders all the iron, alumina, titanic oxide, and phos- 
phorus available for determination. Lime and magnesia, and possibly 
a little baria are not recovered, unless great accuracy is required. The 
amount of lime or baria the residue is likely to contain is negligible, 
TABLE 1 
1 2 3 
SiO, Lae teoe inte ie) CAVING bu alittle 63.54 
(AdnWs”:...> svete eisnrs cute 16.93 0.07 0.4 
5 CH 0 ane ete gear nea ane 1.69 
ee oe ¢ 2.67 ae di 
MODUIN TALL aarau 2.77 0.09 3.3 
CaO. Aker oe: au! 5.07 0.01 0.02 
IR O oe eect A eats nee 4.08 
EON Ore oy ok) ae 2.18 
150), ee eee a ee 0.22 
15 0 era Ree eee 0.04 
INTO Myths Raentod ei Baer babe 0.51 0.16 31.4 
iOpen. case eee 0.14 0.01 On 
Ont h tis. mre, Rate. eee 0.04 
Ste ee REDD Er Cee EN ID 0.02 
IVET Oar. cis. seks cee 0.07 
BaO.. 0.07 
SU ee pe ede eee ie. 100.00 0.52 
1. Andesite. The Crater, Lassen Peak, California. The new lava of 1915. Average 
of one incomplete and two complete analyses. M. Aurousseau, analyst. 
. Oxides of the residue from silica, expressed as percentages of the rock. 
3. Oxides of the residue expressed as percentages of the corresponding oxides of the 
rock. 
to 
and in a rock of the kind studied here the small amount of magnesia 
has but a slight effect on the final result, and its recovery is unim- 
portant for any but the most exacting work. 
The examination of the residue from silica in rock analysis is trou- 
blesome and inconvenient. The amount is small, and must be 
collected specially, if required for study. It is none the less highly 
desirable that residues from a wide range of rock composition (say 
from granite, diorite, basalt, nephelite syenite, and peridotite) be 
examined systematically, in order that some general conclusion 
might be drawn concerning the nature and behavior of the residue. 
