202 Lawson on diabase dykes of the (Rainy lake region. 
and Wadsworth.' Olivine in a similar condition in terrestrial 
rocks has recently been described and figured by Rdnard in 
specimens from Kerguelen island in the Indian ocean.- The 
polysomatic structure in augite is not so well known. R^nard 
notes that the augites of the feldspathic basalt of Heard isla;id, 
Indian ocean, are grouped together at certain points,^ and again 
in the same rocks in Marion island that the augite is characterized 
by a tendency to form groups of individuals having their vertical 
axes parallel.* Teall mentions "granular aggregates" of augite 
in the Hett and the High Green dykes in the north of England.* 
Some of these apj^ear from the figures given to be aggregates 
of grains of augite not in close juxtaposition with an interstitial 
base, although that figured in plate xii, fig. 5, would seem to 
be a polysomatic augite, and if so is the only strictly parallel 
instance that I can find of this structure so common in this dyke 
and in others of the region. 
^ The augite is generally altered to hornblende at its periphery 
and occasionally the latter mineral entirely replaces the former. 
The process of alteration does not appear to proceed along the 
almost or quite imperceptible lines of demarkation between the 
different individuals of the polysomatic augite, but extends from 
the periphery of the mass as a whole in towards its centre. 
The plagioclase appears in two general forms, a rather stout 
or tabular form which is the larger, and usually the more cloudy 
with decomposition products, and a small long lath-shaped feld- 
spar which appears quite fresh and in which the polysynthetic 
lamellae are much more distinct than in the former. 
Magnetite occurs in irregularly bounded masses or is dissem- 
inated, often quite thickly, through the augite as inclusions of 
dusty or finely granular aspect. Pyrite also occurs and is dis- 
cernible macroscojDically. Apatite is seen in occasionally color- 
1 Lithological Studies, Mem. Mus. Comp. Zool. Harvard, vol. x. pi. i. 
2 Notice sur la geologic de 1' ile de Kerguelen, Bui. Mus. Roy. Hist. 
Nat. Belgique, tome iv. No. 4., p. 233, fig. i, pi. v. 
3 Notice sur les roches de 1' ile Heard. Bui. Mus. Roj. Hist. Nat. Bel- 
gique, 1886, 8 p. 260. 
* Notice sur les roches de V ile Marion. Ibid. p. 250. 
5 Petrographical Notes on some north of England Dykes, Q. J. G. S., 
1884, 158. p. 229 and 242. 
