Vol. 58.| THE MINERAL ANALYSIS OF ROCKS. XS 
gravity ; and to this is doubtless due the greatest discrepancy in my 
calculation of the composition of kentallenite, the silica falling 2 per 
cent. below the amount found by analysis. An error of almost the 
same amount appears in two other cases that I have studied— 
one a Tonale diorite from Adamello, and the other a similar 
rock from New Zealand. In the Adamello rock the results of 
a mineral separation and the deduced chemical composition are as 
follows :— 
| | | | 
| SiO, | Al,O,| Fe.O,| FeO | CaO | MgO | K20 | Na.O|} H2O 
| Per cent. |———|———-|——-—_|——_-- |- —__| —— — _— 
| Orthoclase ...... SAU eae PAO} URS eedden. |leoabcant| ooanne.l| seco O54 | 
| QTE SA Re epaeeee2 od Rel erase 22°18 | 
| Andesine ......... 56°62 ...... SOON | V2 SOF | ceseet le scone DOS | sesesen |sesees 3°91 
PB1ObIbe: ...2..2...0- TPESy Skceo 6°45 | 2°54 | 3:07 | 2°95 | 0:29 | 1:22 | 0:98 | 0°03 
| Hornblende...... GG ee oe 2 Beadan noose, || OR) |) Wreko) ) lees} 
Magnetite and) . 
| Bphene. ....;. } ae ea a ol en rs 
64°43 | 16°02 | 3°07 | 3°47 | 4°62 | 2°65 | 152 | 3:94 
Analysis by G. vom Rath...| 66°91 | 15:20 | ...... 6-45 | 3:73 | 2:35 | 0-86 | 3:33 | O16 
| (Total: 98°99.) | | 
Even if we disregard the small quantity of magnetite and sphene, 
it is clear that G. vom Rath’s analysis is not perfect, since it indicates 
all the iron as existing in the ferrous state ; while an independent 
analysis of the biotite, also by him, showed the presence in this 
mineral of a larger quantity of ferric than of ferrous iron. The sum 
moreover is more than 1 per cent. toolow. These inaccuracies, how- 
ever, do not greatly diminish the importance of the differences between 
the results, as found by calculation on the one hand and by analysis 
on the other. The silica in my estimate is over 2 per cent. too low, 
while the alumina, calcium, potassium, and sodium are too high ; and 
this is a natural consequence of an over-estimate of the proportion 
of the ferro-magnesian to the more acid minerals. 
It still remains to point out certain cases in which the mineral 
-separation alone will fail to inform us certainly of the chemical 
composition of the rock. Foremost is the question of the alkalies, 
which may most forcibly be illustrated by an instance such as 
that of the Plauen syenite. Neglecting the quartz, the amount of 
which I have not determined, the mineral composition of this rock 
is as follows :— 
Per cent. 
Orthoclase and microperthite ............... 28°72 
Oligoclase (spec, grav. 2°645) ..............- 51:99 
Hornblende (spec. grav. 3°2) ........0...06+ 16°87 
Zircon, magnetite, and sphene............... 2°42 
100-00 
The chemical composition as deduced from this is as follows :— 
