222 CRUISE of THE NEPTUNE 
the surface of the islands, it would be impossible to map them 
on the scale used in illustrating this report, and in consequence 
the old colouring is followed here. 
CARBONIFEROUS. 
The southern boundary of the Carboniferous sandstones with 
their included coal seams crosses the southern part of Banks 
island in a north-northeast direction, and they consequently 
cover the northern two-thirds of that island, while the extreme 
northwest portion of Victoria island is also occupied by these 
rocks. The western Parry islands on the north side of Melville 
sound are almost wholly formed of these rocks, whose southern 
boundary strikes northeast across the northern half of corn- 
wallis island. They are found again in Grinnell peninsula, the 
northwest portion of north Devon, and again on the western 
side of Ellesmere, in the vicinity of Store Bjornekap, being 
probably largely developed in the northeast part of that great 
island. 
These rocks are described as follows by Professor Haughton : 
'the Upper Silurian limestones, already described, are suc- 
ceeded by a most remarkable series of close-grained, White 
sandstone, containing numerous beds of highly bituminous coal 
and but few marine fossils. In fact the only fossil shell found 
in these beds, as far as I know, in any part of the Arctic Archi- 
pelago is a species of ribbed atrypa, which I believe to be 
identical with the Atrypa fallax of the Carboniferous slate of 
Ireland. These sandstone beds are succeeded by a series of blue 
limestone beds containing an abundance of marine shells, com- 
monly found in all parts of the world where the Carboniferous 
deposits are at all developed. The line of junction of these 
deposits with the Silurian on which they rest is N. E. to E. 
N. E. (true). Like the former, they occur in low flat beds, 
sometimes rising into cliffs, but never reaching the elevation 
attained by the Silurian rocks in Lancaster  
