370 M. Theodor Kjerulf on a 



Or, to afford readier means of comparison with the compo- 

 sition of normal trachyte, it may be calculated at 100, and 

 with protoxide of iron, as follows: — 



Normal Trachyte 

 Silica, .... 76-67 



Alumina, . 



Protoxide of iroi 



Lime, 



Magnesia, 



Potassn, 



Soda, 



} 



;. No. 1. 



' No. 2. 



79-11 



80-81 



11-67 



1017 



1-51 



1-73 



0-47 



0-30 



0-07 



014 



2-93 



4-84 



4-24 



2-01 



14-23 



1-44 

 0-28 



3-20 

 4-18 



100-00 100-00 100-00 



Thus we have very nearly the composition of normal tra- 

 chyte, — the proportion of silica being somewhat larger, espe- 

 cially in the variety in which decomposition has proceeded 

 farthest. 



In order to check the analysis, I endeavoured, with No. 2, 

 which presented most facility for the purpose, to ascertain 

 the per centage of quartz crystals. A certain quantity, after 

 being weighed, was gently crushed, and water being after- 

 wards poured over it, the light supernatant particles of the 

 mass were removed, and the residuum dried and strongly 

 heated : the quartz crystals, which could then be easily dis- 

 tinguished, were separated mechanically from the small red- 

 dish fragments of trachyte, and from a few particles of white 

 quartz. The result was 2*9, or 3 per cent. In this experi- 

 ment, which was very carefully performed, I cannot imagine 

 that there was room for an error of more than one per cent, 

 loss at the utmost, so that the proportion of quartz cannot 

 possibly exceed 4 per cent. 



The question now arises, whether this excess of silica has 

 been conveyed into the rock by infiltration, or whether it has 

 proceeded from a partial elimination of its basic constituents. 

 The answer to this question is, geognostically, of some in- 

 terest. If we take as our basis the composition of normal 

 trachyte,- — and from the local relations of the porphyritic 

 varieties with other trachytic rocks at Kalmanstunga, it 

 seems exceedingly probable that the former have been formed 

 from the latter. We are able, either on the one hand, upon 



