254 J. D. Dana’s Mineralogical Contributions. 
the ratio may be 54: 11 or 1:2, whence would come the formula 
oR St+ BSF or GRe+3R SP ARs SP 
(analogous to that of staurotide) = Silica 23:3, titanic acid 22°9, 
peroxyd of iron 22-4, lime 31:'4=100. It gives 2 per cent. too 
little of silica, according to the analyses, while agreeing closely 
with the results in other respects. 
Mosandrite, in a similar manner, gives for the oxygen ratio of 
the protoxyds, peroxyds and silica 1 :.2 : 3, or of bases and silica 
1: 1 (precisely 16-57 : 15°86); affording the formula 
R3 Si+ 28 Si+-4iH or (¢R* + 3R) Si + 14H, = (Re, BH) Si+ Ag, 
which, excluding the water, is the formula of epidote., The 
crystallization of mosandrite has not been clearly made out. _ 
It is probable, that there are no true silico-titanates or silico- 
tantalates, the titanic or tantalic acid being a base in each case. 
3. Tourmaline. 
The author has shown that the general formula of Tourma- 
ine is (R, #, By Sit 
the oxygen ratio between the silica and all the other bases being 
3:4, as ascertained by Rammelsberg, and this being the only 
constant ratio. The oxygen ratio for the protoxyds, peroxyds, 
and boracie acid, as deduced from Rammelsberg’s analyses, vatles 
greatly. Group I, affords mostly the ratio 4 : 12 : 4,—Group I, 
the ratio 4: 15: 5,—Group II, the ratios 4: 21; 6, 4: 24:7, ete 
—Group IV, the ratios 4: 36: 11, 4: 40; 12, etc.—Group V, the 
ratios 4: 48: 13, 4: 56: 12, etc.* 
For Group I, the special formula is hence 
ke Sif + sRSf4BSi = ks + 2n4+ 1B) sit 
For the Red Tourmaline of Elba, in Group V, ; 
Sty uRStisBSt Aap ian 4 3B) Bit 
These appear to be the extreme variations in the species Tour 
maline, if we exclude the analysis (No. 30) of a somewhat de 
composed variety from Rozena. The formulas for the other gi 
may be easily written in like manner, in accordance with t 
general formula above. 
Axinite has in like manner the general formula 
(Rs, #, B) Si; 
the analyses afford, as the special formula under this type 
R¢ Si+ 28 Si+ 48 Si, stent 
and this formula was suggested by Rammelsberg in his H 
worterbuch, i, 72. 
_* The oxygen ratios deduced by Rammelsberg for the protoxyds, peroxyl 1 
silica, the boracie acid being included with the silica, are for Group }, }* 
I, 1:4:6; Ill, 1:6:8; LV 33 39> 42: Vi. 1.409 725, 
