310 Clarice and Schneider — Experiments xipon the 



most stable of the series. A microscopic examination of the 

 picrolite by Mr. Waldemar Lindgren showed no inclusions 

 which conld account for this difference in behavior, and it is 

 probably due to merely mechanical conditions. 



Upon boiling directly with sodium carbonate solution, none 

 of the serpentines were attacked. By sharp ignition, however, 

 a little silica was sometimes liberated ; which, as in the case of 

 the talc, could be dissolved out and estimated. The quantities 

 of silica thus set free were as follows : 





A 



B 



C 



D 



E 



Per cent 



6-23 



2-00 



2-98 



none 



6-05 





6-34 



2-63 







4-93 



These results cannot easily be interpreted. At most, only 

 about one seventh of the total silica is thus taken out ; and 

 this may represent either impurities in the minerals or secon- 

 dary reactions of an undetermined kind. 



When serpentine is heated to the point of fusion it is split 

 up, as Daubree has shown,* into a mixture of olivine and 

 enstatite. 



H 4 Mg s Si 2 9 - 2H 2 + Mg 2 Si0 4 + MgSi0 3 . 



Upon ordinary ignition, however, this breaking up of the 

 molecule does not always take place ; and when it does occur, 

 it is as a rule only partial. In three experiments the serpen- 

 tines A, C, and D, were strongly ignited over a blast lamp for 

 an hour each, and then treated with strong hydrochloric acid. 

 By this treatment, olivine, if formed, should be decomposed ; 

 while enstatite would remain unattacked in the residue. After 

 evaporation to dryness and extraction with weak acid, the 

 insoluble material was boiled with sodium carbonate solution 

 to remove free silica, washed, dried, and weighed. The resi- 

 dues were as follows : 



A C D 



Per cent 4-39 20-80 39-96 



The residue from the massive Montville serpentine contained 

 43 -28, and that from the chrysotile 41-34 per cent of magnesia, 

 thus agreeing nearly with enstatite in composition. The picro- 

 lite residue was composed of — 



MgO 36-31 



Si6„ 54-88 



R„0„ _ _ 9-26 



100-45 

 * Compt. Rend., hdi, 661, 1866. 



