86 
Proceedings of the Royal Society of Edinburgh. [Sess. 
are marked (a), the latter ( b ). There happen to be 498 of the former and 
502 of the latter. 
We now further calculate for each of these two sets (a) and ( b ) the 
average totals of — 
(2) (Fe 2 0 3 + FeO) + MgO + CaO + Ka 2 0. 
The results are respectively 12'4 and 31*5. These figures enable us to sub- 
divide set (a) and set ( b ) again into rocks respectively below the new 
constant 124 and below the new constant 3T5. We thus obtain the 
following four sets: — 
(a) (a) . 
249 rocks 
(a) (b) . 
249 rocks 
(b) (a) . 
294 rocks 
(b)(b) . 
Total 
208 rocks 
. 1000 rocks 
We now repeat the same process successively by means of the following 
totals : — 
(3) 
MgO + CaO + Na 2 0 + K a O 
W 
CaO + JS r a 2 0 + K 2 0 + A1 2 0 3 
(5) 
Ka 2 0 + K 2 0 + A1 2 0 3 + (Fe 2 0 3 + FeO) 
(6) 
K 2 0 + A1 2 0 3 + (Fe 2 0 3 + FeO) + MgO. 
We thus obtain finally 64 groups, for which the formulae are given on 
Table III., pages 150, 151, and 152. The respective constants are recorded in 
the general table, page 90. The whole of the 1000 rocks are arranged in 
their groups from page 92 to page 149, the final results being given in the 
above-mentioned table, pages 150, 151, and 152. Any other igneous rock 
not included in this list of 1000 rocks may be arranged under its proper 
group by reference to the same constants, page 90. The following examples 
will show this : — 
Rock I. — Felsite porphyry: Knock Mahon, Co. Waterford; recorded 
in the Q. Journal of the Geological Society, year 1900, p. 686, 
sp. grav. = 2'66. 
Rock II. — Alnoite : Naversdile, Orkneys; N. Jahrb., 1902, vol. ii. 
p. 67. 
Rock. 
Si0 2 
H 
O 
to 
H 2 0 
A1 2 0 3 
Fe 2 0 3 
FeO 
MgO 
CaO 
Na 2 0 
K 2 0 
I. 
72-3 
... 
1-8 
9*0 
6*3 
IT 
•0 
1-9 
5-8 
1-5 
II. 
35*5 
20 
6-0 
11*7 
5*9 
6-3 
13-6 
15*8 
1-9 
2-2 
