442 A. P, COLEMAN 
somewhat irregularly shaped and placed. The latter are prob- 
ably hematite and produce the bronze luster seen on basal 
planes of the corundum. Extinction is parallel to the needle- 
like inclusions, and there is a rather strong dichroism, violet 
when the needles are parallel to the chief section of the lower 
nicol and reddish-brown in the opposite position. Some frag- 
ments, no doubt parallel to the basal plane, are not dichroic. 
The first of the two specimens might be named a plagio- 
clasite (anorthosite contains a more basic feldspar) if taken sepa- 
rately, but the second does not differ from typical examples of 
the York branch nepheline-syenite except in the complete 
weathering of its nepheline, and probably both are varying 
forms of the same rock mass. 
Through the kindness of the director of the Bureau of Mines 
specimens of corundum rocks from Raglan township in Renfrew 
county, about twenty miles northeast of Dungannon, have been 
placed at my disposal. One is white, somewhat schistose, and 
much like the Methuen specimens except that it contains biotite, 
and that the pale greenish corundum crystals are almost an inch 
in diameter and have no bronze shimmer on basal planes. 
Under the microscope it is found to consist mainly of plagio- 
clase (oligoclase) and biotite, the latter pale greenish-brown, 
faintly dichroic and with a small axial angle. There are also a 
few large patches of colorless muscovite having a large axial 
angle. The specimen has the mineralogical composition of a 
diorite, though of a very unusual character; but Professor Miller’ 
states that nepheline-syenite occurs close by, apparently part of 
the same rock mass, though not so highly corundiferous.* 
These white rocks were taken for limestone by farmers of 
the region, and an attempt was made to burn them for lime, of 
course, in vain. Hand specimens, partly fused, were taken from 
the kiln and supposed to be nepheline-syenite, many of them 
doubtless having that composition; but the one provided for 
microscopic examination contains no nepheline. It is evidently 
part of a bowlder and is schistose and pale gray to white on the 
*Bur. Mines, Ont., 1897, p. 222. 
