198 F. W. Clarke— TschermaVs Theory of the 



It must be remembered that in order to account for rump- 

 fite, Tschermak assumed the existence of the constituent mole- 

 cule At 7 ', which appears in none of his other chlorites. Both 

 in his scheme and in mine the mineral is exceptional, and 

 undoubtedly it needs further studj T . 



In amesite, which occurs with corundophilite, we find 

 another chlorite of peculiar composition. It is the most basic 

 member of the group, and contains more oxygen relatively 

 to its silicon, than any other chlorite known. Its formula, as 

 written by Tschermak, is Al,Si0 7 , H 2 (MgOfI) 2 , which is the 

 equivalent of his serpentine molecule, with Al 2 in place of 

 MgSi. As we have no experimental evidence from which to 

 reason, we may with equal propriety regard amesite as having 

 the structure 



°<Mf'> Si0 4 = ( A1H A) 2 ; 



when it becomes analogous in structure to the other chlorites, 

 standing in a similar relation to them as that which the clin- 

 tonite group bears toward the micas. On this supposition the 

 micas and chlorites, as has already been shown elsewhere,* 

 form two parallel series of compounds, as follows : 



Normal orthosilicate, Al 4 (Si0 4 ) 3 Mg 4 (Si0 4 ) 2 



Whence derive — Micas Chlorites 



Muscovite, Al 3 (Si0 4 ) 3 R 3 Aphrosiderite, Mg 3 (Si0 4 ) o R„ 



Normal biotite, Al 9 (Si0 4 ),R 9 Orthochlorites, Mg 2 (Si0 4 ) 2 R^ 



« phlogopite, Al 2 (Si0 4 ) 3 R 9 Mg(Si0 4 ) 2 R 6 



Clintonite, A1<°>R" Amesite, 0<K g >Si0 4 = R 2 



\si0 4 = R 3 



In certain respects, this general scheme is not incompatible 

 with Tschermak's views. The orthochlorites, for example, he 

 regards as mixtures of serpentine and amesite, repesenting the 

 structure of those compounds in a special way. But if serpen- 

 tine be written Mg 2 (Si0 4 ) 2 H 3 (MgOH), and amesite as I have 

 just suggested, the theory of substitution is satisfied, and the 

 orthochlorites may be classified as intermediate between the 

 two end terms. Then, however, the leptochlorites fall less 

 easily into line, and the slight gain in apparent simplicity in 

 the ortho series is offset by loss elsewhere. It is quite proba- 

 ble, however, that corundophilite is a mixture containing 

 amesite molecules ; and the formula which I give the latter 

 mineral is sustained to some extent by the fact that both min- 

 erals occur intimately associated with diaspore. At Chester, 

 the only locality for either corundophilite or amesite repre- 

 sented by analyses, the chlorites are commonly found in seams 



* Clarke and Schneider, this Journal, Sept., 1891. 



