Vol. 58. ] THE MINERAL ANALYSIS OF ROCKS. 173 
of fine flour washed away is in the first place inversely proportional 
to the specific gravity, and in the second dependent on the form of 
the grains and the brittleness of the material. Quartz has given 
rather widely differing results, and is still under investigation. 
It appeared possible that, when two different minerals were 
pounded up together, the amount of flour produced might not be 
the same as when they were pounded separately. To determine 
this, an experiment was made with orthoclase and hornblende—two 
minerals which yield very nearly constant results when washed 
separately. Some fragments of adularia, weighing 4°3158 grammes, 
and of basaltic hornblende, weighing 2°6846 gms., were crushed in 
a steel mortar and passed through a fine sieve: the resulting 
powder weighed 6°8290 gms., so that 2:45 per cent. of the material 
was lost in the process of comminution. Of the powder 3°6271 gms. 
were taken for washing, and lost 0°8641 gm. or 24°05 per cent. 
The washed and dried powder was placed in the separator, 
2:7150 gms. being operated upon; it yielded 1°6201 gm. of ortho- 
clase and 1:0931 of hornblende, the loss being 0:0125 gm. Let the 
original weight of the orthoclase be represented by Or and that of 
the hornblende by Hb; also let the loss of orthoclase be represented 
by « and that of hornblende by y. ‘Then, reducing the quantities 
given above to percentages :— 
G) OF _ 6165, Gy Ore _ 5971, 
Fb 38:55 Hb—y 40°29’ 
and 
e+y=24:00. 
Substituting the values of Or and Hb in (ii) it follows that 
Bo 
a 0201 
4°81. 
Hence 
2=16268 or 26°39 per cent. 
7 
Y=T 132 or 20:16 per cent. 
The orthoclase is thus seen to have lost 26°39 per cent. in washing, 
and the hornblende 20°16 per cent. ; but to compare these numbers 
with the loss on separate washing we have the proportion 
20) 2016 = 22565 Wic2o: 
The ratio of loss of orthoclase to hornblende was thus 22°58 : 17-25, 
while with separate washing it was found to be 22°58 : 16°27. 
The hornblende would thus at first sight appear to have lost 1 per 
cent. more in the process of crushing and washing with orthoclase 
than when treated alone. But, to investigate this point further, 
orthoclase and hornblende were next crushed separately, weighed 
quantities of the powders were then mixed and washed, and after- 
wards separated. Without entering into details, it may be sufficient 
