CLARKE.] 
GARNET-BIOTITE TYPE. 
27 
iu some gehlenites obtained from furnace slags the same replacement 
has been observed.* 
At the Orawitza locality, gehlenite occurs in rolled pebbles contain- 
ing grains of vesuvianite. As this gehlenite contains little or no water, 
we may compare Janovsky's analysis of itt with the results obtained 
from calculation, both with the formula Ca 3 Al 2 Si.jO 10 , and with the new 
formula Al,(Si0 4 ) 5 K // 6 (A10 2 E // )2, when R" 8 =Oa 7 Mg. 
The results are as follows : 
Si0 2 . 
A1,0 3 . 
Fe,0 3 
FeO . 
CaO. 
MgO. 
Iffu. . 
T t Old 
Janovsky. fornmla> 
32.39 
18.53 
1.25 
3.61 
37. 65 
6.W 
.51 
30.77 
26.15 
43.08 
New 
formula. 
32.05 
21.80 
11. 
•21 
100. 63 
100. 00 
100.00 
Neither formula compares perfectly with the analysis, but the new 
is rather better than the old. With some other analyses this advan- 
tage is reversed. We need new data, based upon purer and more defi- 
nite material than seems to have been as yet collected. If the old 
formula should be sustained, then it might be written structurally in 
either of two ways, thus : 
X)— Al O 
Al— SiQ 4 ^ 
\ Si o^ Ca * or 0a< 
,Si0 4 -Ca(AlG) 
Si0 4 =Ca(AiO) 
The first formula would serve to connect gehlenite with garnet and the 
micas, while the second represents a derivative of calcium orthosilicate 
and fails to indicate the various relationships of the mineral to other 
species of this group. Neither formula, however, is so suggestive as 
the one which is based upon the type of vesuvianite. 
Another alternative is offered by the theory of Vogt,f who regards 
melilite and gehlenite as varying mixtures of two end products; one, 
the "gehlenite silicate," K /// 2R // 3Si 2 O 10 , and the other, akermannite, 
R'^SiaOm. The last compound is a basic silicate, found in slags, which 
needs further study. Vogt's theory is plausible, and, if established, will 
take the two species in question, gehlenite and melilite, out of the pres- 
ent category as related minerals to vesuvianite and garnet. 
The little-known mineral arctolite is possibly another member of 
* See Hintze, Handbuch der Min., p. 310. 
1 Dana's System of Mineralogy, 6th ed., p. 476. 
J Neues Jahrb. 1892, II, p. 73. 
