University of California. 



[Vol. 2. 



constituents of actinolite, seems to be abnormal, but as a matter of 

 fact many of the more recent and carefully executed analyses show 

 that these bases are commonly present. A similar remark might 

 be made with reference to the water content. That the water here 

 reported is constitutional can not be questioned since hygroscopic 

 water was determined by drying at ioo° and the material experi- 

 enced no appreciable loss between ioo° and 350 . 



In Table II are tabulated the results of the calculations leading 

 to the determination of the formula of the mineral. It will be 

 noted at once that the oxygen is but slightly in excess of that 

 corresponding to a metasilicate ratio. Most of the theories for the 

 constitution of the hornblende group ignore entirely the presence 

 of constitutional water, which fact probably rests ultimately upon 

 the decision of Rammelsberg* that this is only to be regarded as 

 the beginning of an alteration. In the present material, and 

 apparently the same remark would apply to all of the other 

 analyses of actinolite reported in Table I, no indications of such 

 alteration could be detected. 



The fact that water is reported in many other members of the 

 group, and that it has been admitted as an important constituent 

 of the closely-related pyroxene pectolite, would seem to demand 

 that it should be included in any structural theory of the group. 

 Further difficulties in applying the structural theories of Rammels- 

 berg and Tschermak appear with reference to the sesquioxides 

 and alkalies. According to Tschermak's theory of the horn- 

 blende group, the alkalies must be accompanied by equivalent 

 amounts of sesquioxides, hence any attempt to explain the struc- 

 ture of the present mineral as a mixture of the glaucophane 

 molecule and the actinolite molecule would fail, since the alkali 

 content in both analyses exceeds the sesquioxide content. The 

 theory proposed by Clarkef might be applied with a greater degree 

 of success since it admits the equivalence of K, Na, H, AlO, 

 Fe //; 0. Representing this group of symbols by R', the dyad 

 group by R", and by X a mixture of Si 3 8 and Si0 4 in nearly 

 equivalent proportions, the first analysis may be summed up as 



* Poggendorff's Annalen, Band 103, S. 435. 

 | Bull. 125 U. S. Geol. Survey, p. 94. 



