408 Clarke and Schneider — Experiments upon the 



of soda. This last residue, however, was rich in silica, and 

 therefore could not be spinel. The reaction deserves further 

 study ; but the oxidizability of the iron in prochlorite intro- 

 duces elements of uncertainty which would render it very diffi- 

 cult to interpret the results. 



In the case of the leuchtenbergite, little else was done. By 

 means of Thoulet's solution 5*62 per cent of a yellowish garnet 

 were separated from the mineral, which accounts for part, but 

 not all, of the lime found in the analysis. On this species our 

 results are of little value, except as regards the character of the 

 water which it contains, and its comparative behavior towards 

 gaseous HC1. 



Now, in order to discuss the formulae of the three chlorites, 

 we may reject as adventitious the small quantities of water 

 given up at or below 300°. This leaves as essential water in 

 ripidolite, prochlorite, and leuchtenbergite, 12*65, 11*26, and 

 10 "88 per cent respectively. Using these figures for water the 

 analyses give the following molecular ratios. 



Ripidolite. Prochlorite. Leuchtenbergite. 



SiO, -498 '423 -538 



R„0 3 -186 -241 -185 



R'O "855 -727 -858 



H 2 -703 -026 -604 



Hence we have the following empirical formulae : 



Ripidolite 19RA, 86RO, 70H 2 O, 50SiO 2 



Prochlorite 24R 2 3 , 73RO, 63H„0, 43Si0 2 



Leuchtenbergite... 19R 2 3 , 86RO, 60H 2 O, 54Si0 2 



And these, reduced to an orthosilicate basis become 



Ripidolite R"'s 8 R" 86 H 140 (SiO 4 ) 50 O 113 



Prochlorite R'" 4B R" M H lM (SiO J 42 



Leuchtenbergite R'", 8 R" Bi H 110 (SiO J M 



This excess of oxygen over the orthosilicate ratio can only 

 be interpreted as basic hydroxyl ; whence we get 



Ripidolite R'" M R" M H 1T (SiO J i0 (OH) 111 



Prochlorite R'" 48 R" T , (Si0 4 ) 4 „(0H) 1O7 



Leuchtenbergite R'" M R" 8a H fB (SiOJ M (OH), i 



The last of these formulae is vitiated by the fact that the 

 mineral analyzed was impure ; a fact which appears in the low 

 figure for the hydroxyl, which garnet does not contain. Other- 

 wise it is clear that in general leuchtenbergite and ripidolite 

 agree quite nearly with each other. The question now to be 

 answered is, how shall the hydroxyl be apjjortioned between 

 the bases ? 



128 



