92 PEOCEEDINGS OF THE GEOLOGICAL SOCIETY. 



gneisses and schists is that of securing typical specimens. In the 

 search for these we are necessarily limited to districts where the 

 pre-Palseozoic age of the rocks can be proved by their position 

 beneath, and marked discordancy from, some part of the Cambrian 

 or very lowest Ordovician series, or at least by their forming part of 

 a series which can be somewhere seen in this position. But even 

 in these districts the great age of the rocks renders them liable to 

 subsequent alteration either by the percolation of water, or by 

 crushing followed more or less by mineral reconstruction. Thus 

 regions which have been the scene of mountain-making are unfa- 

 vourable to our quest. Hence our own country does not provide us 

 with good types, although its rocks may become useful at a later stage 

 of our investigations. Those which I have studied with most advan- 

 tage are from the Laurentian series of Canada, the antiquity of 

 which is beyond dispute, and from districts which appear to have 

 been singularly free from any marked disturbance since an epoch 

 long anterior, at any rate, to that when the basal materials of the 

 Potsdam Sandstone were deposited. The macroscopic aspect of 

 these rocks has been already described. I will endeavour to 

 indicate their more characteristic microscopic structures, although 

 it is most difficult to describe them in words, without the aid of 

 large coloured figures. 



The gneisses in the Laurentian series — for to these, as the 

 dominant rocks, I will restrict myself — mineralogically do not differ 

 materially from granites : quartz, felspar (often potash), and black 

 mica or hornblende are the principal minerals ; white-mica, garnet, 

 apatite, sphene, magnetite, or some iron-oxide, are rarer and 

 more accidental constituents. They are holocrystalline, presenting 

 many resemblances macroscopically to granites of moderate coarse- 

 ness, but distinguished by a tendency to mineral banding, as above 

 described, which is sometimes very definite. If we examine 

 slices cut perpendicular to the banding, we note that the sections of 

 the mica lie with their longer axes roughly parallel with it. The 

 same remark applies also to any mineral of elongated form ; and 

 minerals of the same species exhibit some tendency to be associated 

 in these layers. Any such grouping is extremely rare in a granite, 

 and the circumstances under which it occurs suggest an explanation. 

 The two rocks, however, agree in that the mica-crystals are well 

 developed, and of a fair thickness compared with their basal area 

 — i. e. are not mere flakes. Again, in a moderately coarse granite 

 (I expressly exclude the very fine-grained granites, which usually 

 occur in veins and small bosses, the rock called granulite by some 

 authors), the crystals of felspar, where in contact with quartz, 

 exhibit, as a rule, a rectilinear boundary; but in these gneisses, while 

 the boundary between the two minerals is clear and definite, 

 it is curvilinear, the grains of both being either rounded or even 

 somewhat lobate in outline, as if the two minerals had simultaneously 

 but slowly segregated out of a viscid mass. Thus, according to 

 the predominance of either mineral, rounded grains of quartz may be 

 seen included in larger grains of felspar or vice versa. The struc- 



