Constitution of certain Micas and CAlorites. 381 



part of a silicate, but not its actual amount. Probably the 

 hydroxyl is first replaced by chlorine, converting the group 

 -MgOH into the chlorhydrin-like group -MgCl. The latter, 

 by continued action of the gaseous acid is then possibly split 

 ofl as MgCl 2 , but with secondary reactions to which the anoma- 

 lous and irregular results may be due. The question is still 

 open, and we hope by further experiments to get at something 

 more conclusive. The problem is, to measure basic hydroxyl 

 by its replacement with chlorine ; and there seems to be no 

 good reason why it should not be solvable. 



By strong aqueous hydrochloric acid the chlorites and the 

 serpentine were easily decomposed ; but by previous ignition 

 they were split up into soluble and insoluble portions. This 

 splitting up is, in the case of serpentine, already well under- 

 stood ; the products being as Daubree has shown, olivine and 

 enstatite. For the waluewite and the chlorites, however, the 

 existing data were scanty, and accordingly new experiments 

 were undertaken. In each case the powdered mineral was 

 ignited over a blast for several hours. It was then digested 

 with strong hydrochloric acid, and after evaporation to dryness 

 and resolution, the residue was filtered off. Then, after dry- 

 ing and ignition, it was boiled with aqueous sodium carbonate 

 to remove the silica which had been released from the soluble 

 silicates, and the final residue was weighed and analyzed. As 

 this process was followed in our first investigation with the 

 Pennsylvania ripidolite, the results are included here for com- 

 parison with the Siberian material. The percentages of insolu- 

 ble residue were as follows : 



Waluewite. 



Clinochlore. 



Leuchtenbergite. 



Bipiclolite. 



Found, 45-01 



17-56 



19-24 



19-74 



Corrected, 43-96 



16.63 



18-05 



18-49 



In the second line the silica found in the residues, being 

 presumably extraneous, is deducted. The composition of the 

 residues is given below. 





Waluewite. 



Clinochlore. 



Leuchtenbergite. 



Ripidolite. 



Si0 2 



2-34 



5-25 



6-16 



6-32 



RA 



71-12 



67-20 



68-52 



67-81 



MgO 



26-75 



27-89 



25-12 



25-67 



CaO 



trace 















100-21 100-34 99-80 99"80 



In these analyses the silica ranges from 093 to 1*25 per cent 

 of the original mineral, and the remainder has quite sharply 

 the composition of spinel. This point was noticed in our 

 former paper, in the case of the West Chester ripidolite ; but 



