188 Canadian Record of Science. 



thosites in Canada, often in enormous quantities, so that it 

 is considered as particularly characteristic of them, while 

 in the Laurentian proper, the iron ores in the greater num- 

 ber of cases, contain no titanic acid. Lacroix, 1 who has 

 investigated somewhat similar inclusion, s which, however, 

 are double refracting, in certain Norwegian gabbros, thinks 

 that they are pyroxenes, especially as they frequently 

 appear to be grouped together, forming larger grains which 

 may be determined as belonging to this species. He says : 

 " Les grains en question semblent avoir attire* a eux les par- 

 ticules pyroxe*niques en suspension dans le feldspath et les 

 avoir incorpor^es a leur masse." It is quite possible that 

 these inclusions so often found in gabbros and allied rocks 

 consist of the heavier minerals of the rock, in some cases 

 pyroxene and in others iron ore, which were finely dis- 

 seminated through the magma while the rock was crsytal- 

 lizing, or which, perhaps, separated out as the several con- 

 stituents crystallized. My best thanks are due to Professor 

 Judd for a small collection of thin sections of typical gab- 

 bros and peridotites from the north of Scotland which he 

 has described and on which he has principally established 

 his theory of "schillerization." An examination of these 

 shows that nowhere in them are the inclusions in ques- 

 tion so numerous and well defined as in the Canadian 

 anorthosites. The peculiar arrangement of these inclu- 

 sions in the Scotch rocks along cracks, fissures, etc., 

 which Professor Judd has described and which especially 

 supports his theory of their secondary origin, is not ob- 

 served in these Canadian rocks. Their inclusions are on the 

 contrary distributed thickly and pretty uniformly through 

 the whole feldspar individual, generally indeed through- 

 out the feldspar of the whole rock. They disappear as above 

 mentioned only when it has the peculiar granulated char- 

 acter. This remarkable fact will be referred to again. 



The uniform distribution of these inclusions does not 

 prove that they are not schillerization products, for even 



1 Lacroix, Contributions a l'ltude des gneiss a Pyroxene, p. 141. Bull. Soc. 

 Min. Fr. Avril 1889. 



