348 J. H. H. Corbin on certain compounds 
metal. The process of Boussingault, already referred to, was 
used for the determination of the carbon. Alt is pro- 
cess worked satisfactorily with Boussingault, whose compound 
contained a small percentage of chromium, it failed with me 
to fully decompose this specimen, which contains more than 26 
per cent of that metal. 
The carbon was weighed as €0, ina potash bulb, The 
compound yielded by the analysis : 
if 2. 
SEO cage ig Ae are ea 70°18* 70°12 70°23 
Chromium ..-.......-26°25 at a egiay 
FL eae 26 page Pigs t 
Calcium 2°16 ran ited 
Magnesium .-..--..-- 58 <a — 
on trace Se es ee 
C ee ee eee 1°69 ees weer 
FRospbhorag as s.- ~n0 none Pi catin ee 
Sulphur none ee ee 
101°12 
By eliminating carbon and supposing the small percentage 
of other metals than iron and chromium to be replaced by iron, 
the analysis yields almost exactly the formula Fe, €r, ; whic 
is intermediate between 2 and 3 of Smith’s compounds alluded 
to in the beginning of this article. Since the proportion of 
carbon in white cast iron varies from 3°5 to 5°75 per cent, this 
< tenia is peculiar in having a much lower percentage of that 
t. 
shape of an ellipse until the ends touched: on examination no | 
injury to the steel could be detected, as it appeared to possess 7 
the same pre after as before the operation. The steel is ; 
2 The hardness of a tempered specimen 18 se base 
that of quartz, and that of an untempered specimen to that © 
feldspar. The density of the tempered specimen is 779 and 
that of the untempered one 7:8. The density of the latter, 
* Mean of 2 and 3. © 
