Constitution of the Natural Silicates. 307 



Against Groth's formula the stability of talc towards acids 

 tells very strongly. The univalent group — Mg — OH ought 

 to be soluble in hydrochloric acid ; and evidence, to be pre- 

 sented later, goes to show that that particular group is easily 

 removable by the dry, gaseous HC1. In fact, our experiments 

 make it highly probable that all the magnesia taken from a sil- 

 icate by perfectly dry HC1, was originally present in the 

 hydro xylated form. As regards the constitution of talc, how- 

 ever, other evidence is even stronger. 



If Groth's formula is correct, then, upon ignition, talc should 

 behave according to the equation 



Mff <- Si 2 -Mg-OH _ H Q _ M Si 2 -M g> 

 M S< Si 2 — Mg— OH 1X - ~ M S<-Si 2 — Mg-> u ' 



In other words the loss of water should produce but little 

 change, and no silica should be liberated. If, on the other 

 hand, the acid metasilicate formula is true, talc should split up 

 into 3MgSi0 3 + Si0 2 + H 2 ; that is, one-fourth of the silica 

 should be set free ; which, in the }3resent instance, would 

 amount to 15 - 57 per cent. This actually happens ; and a 

 weighed quantity of talc, ignited very intensely for half an 

 hour over a blast-lamp, gave up 15*36 per cent of Si0 2 upon 

 subsequent boiling with a solution of sodium carbonate. Upon 

 longer ignition, as might be expected, a part of this silica re- 

 verts to the insoluble form, and somewhat lower results are 

 obtained. Upon the unignited talc, boiling with soda solution 

 for twenty-four hours produced little or no effect ; soda was not 

 taken up, nor was silica removed. 



In brief, both lines of evidence, the liberation of silica and 

 the stability toward acids, confirm the ordinary formula for 

 talc and controvert the views of Groth ; and no other formula 

 out of several which are possible, satisfies both of the experi- 

 mentally established conditions. The mineral, therefore, must 

 be regarded as an acid metasilicate ; although its ultimate 

 structural formula can be written only when we have a bet- 

 ter knowledge of metasilicic acid. It is a noteworthy fact, 

 that no normal metasilicic ether is yet certainly known ; for 

 Ebelmen's results have not been confirmed by later observers. 

 . Troost and Hautefeuille, however, prepared an ether having 

 the composition (C 2 H 6 ) 8 Si 4 12 . Possibly enstatite may be the 

 normal magnesium salt corresponding to this ether; in which 

 case talc would be a substitution derivative having H 2 in place 

 of one atom of Mg. 



3. Serpentine. 



On account of the importance of this species, its variability 

 in externa] characteristics, and its manifold relations to other 

 minerals, several distinct samples were investigated. 



