184 
In this caleulation certain assumptions have been made. It 
has been assumed that albite and orthoclase are present in equal 
amounts. This is very nearly correct, and a quantitative 
separation of the felspars would have very nearly doubled the 
time occupied in measurement. The variation in the amounts of 
Si O, and Al, O, from this cause is certainly negligible. The 
total amount of alkalis is approximately correct, too. Albite has 
been calculated as pure Ab, whereas it is Ab,, An,. Some 
Ca O is thus neglected, and the Na, O is correspondingly high. 
Ideal compositions have been assumed for orthoclase and 
muscovite. The composition assumed for biotite is that given by 
Messrs. Cross, Iddings, Pirsson, and Washington, in column C, 
Table xiv., of their “Quantitative Classification of Igneous 
Rocks.”* This constituent occurs in such small amount that the 
error involved in using the assumed composition is certainly 
very small. 
Na, О and СаО, which occur in proportions amounting to 
only a fraction of a per cent., have been neglected. 
Iron ores have been calculated as Fe, O,. 
The chief probable errors in the calculated analysis are there- 
fore as follows :— 
Some Ca O is certainly present, but the amount is likely to be 
less than 1 per cent. Na, O is proportionately high. 
Ti O, islikely to bo a very little low, as the Ti O, content 
of the ores has been neglected. 
H, O is sure to be somewhat too low— probably not much. 
Traces of S, CO,, and P, O, must occur. Analyses of two 
other granites are inserted for comparíson. Both are taken from 
Dr. Washington's list of analyses :— 
per TIS TD, 
SEO io TET 69-28 70:65 68:87 
Ads Osaka. |, 16:28 16:16 16:62 
esc E 1:64 1.58 0:43 
Ке О ASI 1:52 0:52 272 
МЕО "m 0:70 trace 160 
СаО "vU mot estimated 0:00 0:71 
Na, O (NUIT 2:90 0 54 1:80 
К. Оо санаса 6:54 8:66 6:48 
Hose v 097 1:20 0:74 
ШО on, меті 0:18 — — 
Б, Өшкен of I au 0:05 
Total a 100:01 99:83 100:02 
* Loc. cit., Table xiv. | Analysis C; 
t N.B.--Decimals are untrustworthy in I. They have been 
introduced to show that there has been no clerical error in calcu- 
lation. 
