662 REPORT— 1897. 



In the second column the water is neglected and the percentage composition of 

 the remaining substances indicated. The analysis confirms the view that a great 

 concentration of iron oxide has taken place, and suggests the further conclusion 

 that there has been a concentration of magnesia and a reduction of the lime, silica, 

 and alumina, thus agreeing with the residts of the microscopic examination. 



Several observers are quoted by the author as having established the fact that 

 magnetite is not always one of the earliest minerals to form, and in basalts of the 

 Franz Josef Land type there is clear evidence that a basic magma may consolidate 

 without any separation of this mineral, although the mother-liquor may contain 

 30 per cent, of iron oxides. 



Brogger, Vogt, and others have observed a tendency in certain dykes for the 

 molecular groups, of which the first-formed mmerals are built to migrate towards 

 the cooling margins. _ The cases examined are mostly those of intermediate rocks, 

 in which the basic minerals are the first to form, so that the margins are more basic 

 than the central parts. But it appears probable that cases occur in which the 

 opposite is true. If the magma of the Franz Josef Land basalts had cooled slowly 

 in a fissure, we should expect to find the central portion of the dyke richer in iron 

 oxide than the margin. Professor Lawson has described two basic dykes from the 

 Eainy Lake region where this is actually the case, and a more striking illustration 

 is seen in the Taberg iron-ore mass, described by Sjogren and Tornebohm, where 

 the marginal portion of an eruptive mass about one square kilometre in area is formed 

 of olivine-hyperite containing only small quantities of magnetite and olivine, 

 which passes mward by gradual stages into a magnetite-olivinite without plagioclase. 

 In conclusion, it is asked whether the metallic iron which occurs as interstitial 

 matter in some of the Greenland basalts may not have been formed by the reduc- 

 tion, by included organic matter, of the iron oxides previously concentrated by 

 progressive crystallisation. 



6. The Glaciation of North-Central Canada. By J. B. Tyrrell. 



In the region immediately west of Hudson Bay the earliest glaciation, of which 

 any traces were recognised, flowed outwards from a gathering-ground which lay 

 north or north-west of Doobaunt Lake. Subsequently this gathering-ground 

 moved south-eastward, until it centred over the country between Doobaunt and 

 Yath-kyed Lakes. From one or other of these centres the ice seems, to the writer, 

 to have flowed westward and south-westward to within a short distance of the 

 base of the Rocky Mountains ; southward, for more than 1,600 miles to the States 

 of Iowa and Illinois ; eastward into the basin of Hudson Bay ; and northward into 

 the Arctic Ocean. 



No evidence M^as discovered of any great elevation of this central area in 

 Glacial, or immediately pre-Glacial, times, and, in the absence of such evidence, it 

 would seem not improbable that the land then stood at about the same height 

 above the sea as it stands at present. In this case the moisture giving rise to 

 the immense precipitation of snow would have been derived from the adjacent 

 waters of Hudson Bay and the Arctic Ocean. 



The name Keewatin glacier has been applied to this central continental ice- 

 sheet. In general character it appears to have been somewhat similar to the great 

 glacier of North-Western Europe, with a centre lying near the sea-coast, a steep 

 and short slope seaward, and a very much longer and more gentle slope towards 

 the interior of the continent. But there was this difierence between the two, that 

 the centre of the latter was over a high rocky country, from which the ice naturally 

 flowed outwards towards the surrounding lower country; while the centre of 

 the former was over what is now, and was probably also then, a low-lying plain, 

 on which the snow accumulated to such depths as to cause it to flow over country 

 very considerably higher. 



After the Keewatin glacier had reached its full extent, it began gradually to 

 decrease in size. As it disappeared from the Northern States, and the North- West 

 Territories of Canada, it left a series of moraines, many of which can be readily 

 traced across the unwooded country, as ridges of rounded stony hiUs. "UTiile 



