Intelligence and Miscellanies. 197 



after which they weighed 1 1 .62 grs. The mixture was now 

 digested with muriatic acid until the oxide of iron was whol- 

 ly dissolved ; the silica remained behind in white flocks, and 

 was separated by the double filter, washed, dried and ig- 

 nited. Its weight was 7.53 grs. This amount, deducted 

 from 11.6-2 grs. gives 4.09 grs. per oxide of iron, which, re- 

 duced by calculation to the protoxide, the condition in which 

 it probably exists in the mineral, equals 3.68 grs. 



The constituents of this mineral therefore appeared to be, 

 in this instance, 



Silex, 7.53. 



Magnesia, ..... 5.50. 



Protoxide of iron, . . . 3.68. 



Soda, . . ^ 



Oxide of chrome, > . . . 1.09. 



100.00. 



Considering the soda and oxide of chrome as accidental, 

 the preceding analysis, it will be observed, agrees very well 

 with the supposition that the present variety of chrysolite is 

 a compound of one atom bisilicate of iron, with three atoms 

 silicate of magnesia ; and its coincidence with the miner- 

 alogical formula /S^ -f- 3MS will be still more striking, if 

 we suppose the oxigen of the iron is estimated a little too 

 high, in consequence of the probable union of a small por- 

 tion of that metal with sulphur, to form the proto-sulphuret 

 of iron, — a substance whose mechanical admixture, in a slight 

 degree, with this mineral was sufficiently evinced by our first 

 experiments. 



I am aware that the difference in composition between 

 the specimens just examined, and those of the chrysolite 

 analyzed by Klaproth and Stromeyer* may seem opposed to 



* The specimens examined by Klaproth came from the Levant ; the formu- 

 la of whose composition is/S-j- 4MS. For Stromeyer's analysis of meteoric 

 chrysolite, see Vol. 13, p. 184, this Journal. 



