290 ME. W. R. WATT 0~S THE GEOLOGY [June 1914, 



Frequent reference has been made in the descriptions of the 

 cordierite-norites to the prevalent association of hypersthene, 

 magnetite, and colourless garnet. It may be useful to recall the 

 following facts : — 



(1) The colourless garnet bears as inclusions particles of magnetite. 



(2) It is usually accompanied by pink garnet, into which it passes 

 insensibly. 



(3) It possesses vermicular markings, which are likewise found on the 

 edges of the pink garnet. 



(4) The colourless garnet persists so long as there is hypersthene 

 present, and when no hypersthene is present, pink garnet alone occurs. 



(5) The colourless garnet, when surrounding hypersthene or magnetite, 

 may or may not be developed towards felspar. 



(6) With the formation of pink garnet, there seems to ensue a 

 diminution in the amount of magnetite included in the garnet. 



From these facts it appears that the hypersthene is being 

 destroyed to form, by interaction with the anorthite-molecule of 

 the felspar, garnet and a less basic felspar. Magnesia would seem 

 to be lost in the process — the introduction of potash from external 

 sources, however, fixes this as biotite. In the change from hyper- 

 sthene to ordinary garnet, the colourless variety with inclusions of 

 magnetite constitutes an intermediate stage. 



In regard to such garnet-rims, the following previous notices 

 may be of interest. In a gabbro from the Philadelphia district, 

 Miss F. Bascom 1 describes reaction -rims surrounding pyroxene. 

 Usually the pyroxene is hypersthene, but augite or diallage 

 accompany or replace it. The plagioclase is labradorite or 

 labradorite - by townite. Whether rhombic or monoclinic, the 

 pyroxene is surrounded by a reactionary peripheral rim consisting 

 on the outside of garnet, and within of quartz and hornblende. 



In a paper on the origin and growth of garnets, Sir Thomas 

 Holland 2 describes reaction-rims, composed of two layers, between 

 hypersthene and garnets. Next the garnet comes a layer of 

 vermiform pieces of green actinolite, together with felspar showing 

 lamellar twinning. Next the hypersthene is a zone composed of 

 magnetite-granules. In these rocks also, the hypersthene may be 

 accompanied by augite, hornblende, biotite, graphite, and garnet, and 



* sometimes it is replaced entirely by garnet in rocks which . . . are beyond 

 question altered forms of the pyroxenic series.' (Op. cit. p. 23.) 



The centre of the garnets is pink, but towards the margin they 

 show the vermiform cavities with which the above-mentioned 

 actinolites communicate. Along this zone the change proceeds, 

 garnet substance being laid down in the cavities. In this case 

 the garnet grows towards the hypersthene, and not towards the 

 felspar as in the Pennsylvanian case. It is also suggested that 

 there is some relation between the depth of colour in the garnet 

 and the original hypersthene. There is thus a striking similarity 



1 Bull. Gool. Soc. Am. vol. xvi (1905) p. 313. 



2 Rec. Geol. Surv. India, vol. xxix (1896) pp. 20-30. 



