

210 HOLLAND: S1VAMALAI SERIES. 



alumina present to use tip the magnesia as spinel. 1 Morozewicz 

 is repeatedly explicit on these two points, and I consequently do not 

 see in what sense his experiments can offer any support to 

 Dr. Pratt's conclusions as to the origin of the corundum associated 

 with the peridotites in North Carolina. At the same time, I am quite 

 conscious of the fact that in Nature the physical conditions may so 

 far depart from those under which laboratory experiments are con- 

 ducted that it may be possible for corundum to separate as a primary 

 constituent in a dunite magma. Where, however, the conditions 

 of a natural occurrence so markedly differ from those which are 

 considered to be essential in a laboratory experiment, it is necessary 

 to be all the more careful to be quite sure that the natural condi- 

 tions have been correctly interpreted. Dr. Pratt's description of 

 the occurrences in North Carolina, if I have not misunderstood it, 

 is suggestive as much of the formation of corundum by contact 

 action as by the separation of free alumina through differentiation 

 of a molten magma, and in view of Morozewicz' s work the latter 

 conclusion requires more than ordinary criticism. There was a 

 time when we doubted the occurrence of corundum as a primary 

 constituent of igneous rocks. Let us be careful now not to commit 

 the converse error, and regard it only as a pyrogenetic primary 

 mineral : it is a rare mineral indeed that can be formed by one 

 only of the ways in which we imperfectly classify the processes of 

 Nature. 



Morozewicz has given a description of an occurrence in the 

 Urals of corundum in felspar-rock very similar to the instance at 

 Sivamalai. Like the latter, also, the rock occurs both in the pegmatitic 



Corundum-syenite in and the ordinary form of medium texture. The 



the Urals. pegmatite (" corundum-pegmatite ") is composed 



of blue corundum, microperthitic orthoclase, some secondary mus- 



covite and accessory rutile, apatite, zircon and a black, highly 



1 See Zeitschr.f. Kryst., XXIV, 1895,281—285, and op, cit., pp. 34, 56 and 

 72. 



( 42 ) 



