on Serpentines in the Pennine Alps. 539 



a similar mineral in rocks of this character ; it has a general 

 resemblance to zoisite, but I cannot venture to identify it. 



I am indebted to the kindness of Mr. J. C. Chorley for a bulk 

 analysis of this rock (made in Professor Ramsay's laboratory at 

 University College). It is as follows (No. I.) : 



No. II. 1 



28-56 



39-54 



0-99 



2-87 



15-79 



1-73 



0-70 



11-09 



100-35 101-27 



Specific gravity of the rock 2-81. This result appeared to be of so 

 much interest that Mr. Chorley kindly verified it by a partial analysis 

 of another sample, and obtained SiOy=29'88, Al 2 3 =43-56, Fe 2 O a = 

 10 - 65. Thus there can be no doubt that the rock, and consequently 

 its chief constituent, has an unusually low percentage of silica and 

 an unusually high one of alumina. It bears macroscopically and 

 microscopically a very close resemblance to a rock described by 

 myself from near Rhoscolyn, Anglesey, and an analysis of that rock 

 made by Mr. F. T. S. Houghton is quoted for comparison (No. II.) ~ 

 The analyses present considerable resemblances : the ratio of the 

 Si0 2 to the A1 2 3 is but slightly different, there is a larger amount 

 of water in the Anglesey rock, and a smaller amount of the iron 

 oxides. This, however, may be explained. The first is strictly a 

 bulk analysis, not so the second. This rock contains rather con- 

 spicuous octahedral crystals of magnetite (probably containing some 

 Cr 2 3 ). These were removed because the analysis was made to 

 ascertain of what mineral the matrix was composed, and what was 

 its history. The most marked discrepancy between the two analyses 

 is the amount of MgO, which indicates the presence, in the Rhoscolyn 

 rock, of a mineral richer in magnesia. For purposes of comparison 

 let us denote the alumina in each rock by 100. Then we have 

 approximately. 3 



No. I. (Gorner Grat). No. II. (Rhoscolyn). 



SiOo = 65 72 



FeO = 6-5 7 



MgO = 9 39 



Hence it seems to follow that the latter rock must either consist of 

 a somewhat different mineral or contain a second mineral rich in 

 magnesia. Still both analyses seem to indicate the presence of a 

 mineral with a low silica percentage, a moderate one of magnesia and 

 iron-protoxide, and a very high one of alumina. 



1 Also traces of Cr 3 3 and MnO. 



2 Q.J.G.S. vol. xxxvii. (1881) pp. 43-45. 



3 The Fe2()3 is not indicated m the comparison, because I believe most of it is 

 present in JNo. I. as hematite. Thus it will not help us for the present purpose. 



