Washington — Igneous Rocks from Eastern Siberia. Yll, 



As the analysis showed an excess of alumina amounting to 

 three per cent, it was thought possible that corundum might 

 be present. The rock powder was accordingly digested with 

 hydrofluoric acid and the residue examined. None of this 

 mineral was found, however, the only substances unattacked 

 being fluorite (whose identity was thus surely established^ 

 zircon, and a little of the hornblende. 



An analysis of this rock yielded the results given in I, in 

 the other columns being given analyses of rocks of similar 

 chemical composition from other localities. 



It is seen that this foyaite differs decidedly from nearly all 

 others in the ratio of K 2 : Na 2 0, this being close to unity, 

 while in most of these rocks the soda predominates over potash 

 to a very considerable extent, i. e. over 1*5 : 1. Indeed, the 

 Beemerville rock, whose analysis is given above, is almost the 

 only one of this group which approaches the Siberian in this 

 respect. 



The close similarity of the Italian leucite-phonolite, which 

 differs only in slightly higher ferric oxide and lime, is very 

 interesting, as another example to be added to the rapidly 

 increasing list of rocks which, with similar chemical composi- 

 tion, differ radically mineralogically. The difference here is 

 of course to be connected with the difference in geological 

 occurrence ; the Siberian and New Jersey rocks being intru- 

 sive, while the Italian is a surface flow. This tendency of 

 leucite to form in extrusive rocks, while it is absent in intru- 

 sives of the same chemical composition, is a well known fact, 

 but the present case is so well marked, and the similarity in 

 chemical composition so great, that it seems worth while call- 

 ing attention to it. 



As to the mineral composition, if the attempt be made to 

 calculate hydronephelite using the extra alumina as a basis, we 

 get 24*1 per cent of this, 4*7 of albite and only 4*5 of nephe- 

 lite. This result does not correspond with the appearance in 

 thin section, which calls for much more nephelite. I have, 

 therefore, calculated the rock as free from hydronephelite, 

 which is an alteration product, the excess of alumina remain- 

 ing uncombined. As has been pointed out elsewhere, this is a 

 marked feature of other rocks belonging to the foyaite group* 

 The result is given in la below, the composition of the Brac- 

 ciano leucite-phonolite being given in Ilia. 



The albite molecule in la is higher than the appearance of 

 the orthoclase would indicate, as has been said above. It must 

 be noted, however, that the hydronephelite present would 

 reduce the amount of this considerably, as already seen. 



*H. S. Washington, Jour. Geol., vol. ix r p. 609, 1901. 



