MI N ERA L TRA NSFORMA TIONS. 28 1 



its relationship to the Micas, regarded the mineral as a 

 molecular compound Mg 6 (Si G 4 ) 3 . H 4 Si 5 I2 , the two com- 

 ponents of which are according to his theory constituents 

 of the Mica group. 



(3) Serpentine H 4 Mg 3 Si,0 9 is the almost universal result 

 of the decomposition of Olivine Mg 2 Si 4 . 



Scharizer regards both these minerals as di-orthosili- 

 cates ; Olivine is then ( Mg„0),Mg 4 (SL0 7 )„ and Serpentine 

 H 2 (MgOH) 2 Mg(Si 2 7 ). 



Clarke regards both minerals as orthosilicates : Olivine 

 as Mg 4 (Si0 4 ) 2 and Serpentine as (Mg OH)H 3 Mg 2 (Si 4 ), 

 This formula for Serpentine was suggested by the results 

 of Clarke and Schneider's experiments with dry hydro- 

 chloric acid upon the mineral ; they found that only one of 

 the three atoms of magnesium is readily extracted, while 

 the remaining two offer greater resistance to the action of 

 the o'as. 



Brauns {Neiies [ahrb. f. Mineralogie (1), p. 205, 1894) 

 has attributed this difference to the presence of water ; this 

 causes the gas at first to act energetically upon the mineral 

 which is then protected by the chloride formed. Experi- 

 ments made by Lindner have also failed to confirm the 

 results of Clarke and Schneider. 



BraunS himself agrees with Scharizer in attributing to 

 Serpentine the formula H 2 (Mg OH) 2 MgSi 2 7 , which can 

 be derived by the removal of magnesium and accompanying 

 hydration from Mg 4 (Si 4 ) 2 or Olivine, after passing through 

 an intervening stage in which the composition is H(Mg OH) 

 Mg 2 Si 2 O r 



Two other facts must be borne in mind in devising a 

 probable formula for Serpentine : firstly, that when it is 

 heated the water is liberated at different temperatures, one 

 molecule going at red-heat, the other at a higher tempera- 

 ture ; secondly, that when Serpentine is fused it becomes 

 converted into a mixture of Enstatite, Mg Si 3 , and Olivine, 

 Mg 2 Si 4 . Both these facts are more or less indicated by 

 the formula just given. 



The graphic formulae used by the above-mentioned 

 authors in support of their various views have not been 



