Chlorite Group and its Alternative. 195 



In short, these chlorites are apparently compounds of pre- 

 cisely the same type as the orthochlorites. Daphnite and 

 chamosite vary from theory by small amounts in opposite 

 directions, the one having a trifling excess of an FeOH salt, 

 the other an excess of an A1H 2 2 compound. Closer agree- 

 ments could not be reasonably expected. 



In order to properly interpret the remaining leptochlorites, 

 we must return to the fundamental idea of the substitution 

 hypothesis. Starting with the normal ortho-salt, we get at 

 once the following probable series of derivatives. 



Olivine, Mg 4 (3i0 4 ) 2 



Mg 3 (Si0 4 ) 2 R' 2 



Orthochlorites, M^(SiOJ 2 R' 4 

 Mg(Si0 4 ) 2 R' 6 



In the first of these derivatives, the compound intermediate 

 between the orthochlorite type and olivine, we find the key to 

 most of the leptochlorites. Again discussing the analyses 

 mainly considered by Tschermak, omitting strigovite and 

 rumpfite for separate inspection, we get the subjoined semi- 

 empirical expressions : 



Metachlorite, Zeynek, Fe" 46 (SiO 4 ) 40 . H lo A 4l M„ 



Earalite, Wiik, Mg M (Si0 4 ) M . H 69 A 32 M 4 



Diabantite, Hawes, Mg fi9 (Si0 4 ) 66 . H M A SB M 9 



Klementite, Klement, Mg 91 (Si0 4 ) 48 . H 7 A M M, 



Delessite, Heddle,* Mg 76 (Si0 4 )„ . H 89 A S8 M 1 + 34 aq. 



Cronstedtite, Ludwig, Fe 49 (Si0 4 ) 35 . A 44 , nearly. 



Aphrosiderite, Raniraelsberg, Fe 64 (Si0 4 ) 41 . H 3 M 8 A 44 



Metachlorite, at one end of this series, corresponds nearly to 

 the orthochlorites ; while aphrosiderite, at the other end, ap- 

 proaches the composition Fe 3 (Si0 4 ) 2 A 4 . All the other min- 

 erals in the series fall between these extremes. Reduced to 

 general form, the expressions become : 



Metachlorite, 4R" 2 (Si0 4 ) 2 R' 4 + lR /, 3 (Si0 4 ) 2 A 2 



Euralite, 



Diabantite, 



Klementite, 



Delessite, 



Cronstedtite, 



Aphrosiderite, 



These ratios, of course, are not rigidly exact, but they vary 

 no more from the analyses considered than the analyses of the 

 different occurrences in any one species vary among them- 

 selves. In the last case, the actual analyses show small admix- 

 tures of the orthochlorite type ; but the approximation is 



* The delessite from Bowling Quarry. The samples from Dumbuck and Long 

 Craig agree well with this. 



7 



u 



+ 3 



4 



u 



+ 3 



1 



a 



+ 1 



1 



a 



+ 2 



1 



u 



+ 4 







u 



+ 1 



