464 SCIENCE PROGRESS. 



hypersthene, quartz, and garnet ; or again a zone of brown 

 hornblende followed by one of garnet. While speaking 

 doubtfully of the hornblende and biotite, Kemp considers 

 the garnet certainly secondary on account of its peculiar 

 relation to the felspar. The quartz he regards as the 

 residual silica liberated in the conversion of labradorite to 

 garnet. In view of the great variety of micrographic inter- 

 growths known to occur as original products in various 

 igneous rocks, the secondary origin even of the garnet in 

 this case seems to be by no means conclusively established. 

 The hypothesis of secondary reactions is confronted in some 

 cases by chemical difficulties ; thus, it is not easy to see 

 how brown hornblende can be produced by reactions 

 between auoqte and labradorite, or biotite from magnetite 

 and labradorite. It is, however, true in a general sense, 

 as has often been pointed out, that the rims are usually 

 more or less intermediate in chemical composition between 

 the minerals which they separate. Allowing due weight to 

 this fact, it is still to be observed that the succession of 

 zones is in general what would be expected if the minerals 

 had crystallised out from a molten rock-magma, following 

 the normal order of consolidation of the several minerals 

 as laid down by Rosenbusch. It is also a common feature 

 in undoubted products from igneous fusion that the earliest 

 formed minerals tend to serve as a nucleus for the crystallisa- 

 tion of those that follow, and this is most noticeable in 

 plutonic rocks of basic composition. Further the "celyphite" 

 borders round garnets, and other similar phenomena, afford 

 numerous examples of radiating fibrous arrangement, and 

 linear parallel intergrowth of different minerals, which in 

 many cases are, beyond reasonable doubt, of primary origin. 

 It seems, therefore, that we must require more convincing 

 proof before accepting the view that the "reaction-rims" in 

 these gabbro-rocks are in general, or in any considerable 

 part, really due to secondary reactions between the original 

 constituent minerals. 



Another interesting feature in these American gabbros 

 is the occurrence in them of considerable bodies of iron-ore, 

 which have clearly been secreted from the gabbro-magma 



