Sei'tember 15, 1910] 



NATURE 



335 



at right angles and that of the Appalachians at an acute 

 angle. 



Of what kind were the mountains erected on these 

 bubblo-like foundations of gneiss, set in meshes of schist? 

 In many places they do not seem to have formed con- 

 tinuous ranges such as those of the Rockies, but rather 

 groups of domes of various sizes. Some of them were 

 comparatively low ; others seem to have been lofly, though 

 broad. Of the low ones, the best known is that of the 

 Grande Presqu' Isle in the Lake-of-the-\Voods, an oval of 

 gneiss eighteen by thirty-two miles in dimensions. Here 

 the up-swelling could not have been great, since the schists 

 dip away from the gneiss at low angles all round, and 

 patches of green schist, remnants of the roof, or perhaps 

 of unusually large blocks sloped from above, are found 

 here and there in the interior. 



. On the other hand, the Rainy Lake batholith, thirty In- 

 fifty miles in dimensions, must have risen as a lofty 

 dome, since the surrounding schists dip away at high 

 angles (60° to 90°). The arch of whicli they were the 

 bases must have swung thousands of feet above the pre- 

 sent surface of the batliolith. Passing inw'ards from the 

 Keewatin, one finds at first immense slabs of the schist 

 shifted a little and enclosed in gneiss, then bands of green 

 material with softened edges, and finally darker cloudy 

 strealis in the gneiss, representing more perfectly digested 

 bands. As Lawson has shown, the outer edge of the 

 batholith is of greyish hornblende syenite gneiss or horn- 

 blende granite gneiss, w-hile the interior is of ordinary 

 mica granite gneiss. The outer part has absorbed a 

 certain amount of basic Keewatin material. 



One cannot doubt that this zone of green schist frag- 

 ments, followed by greyish hornblende roclc, originally 

 extended over the dome as well as round its edges. In 

 the middle there is now a width of ten or twelve miles 

 of the ordinary Laurentian gneiss. This implies, of course, 

 that the upper part of the dome, afterwards removed, was 

 several miles in thicl-iness, and that thS mountain mass 

 rose correspondingly above the synclinal valleys. It must 

 not be assumed that the dome had a regular surface, nor 

 that it was unbroken. Such a batholith as that of Rainy 

 Lake was not made by a single sudden up-welling of 

 granite, but by a long succession of slow inflows from 

 various quarters. Meantime, the rocks above must have 

 been stretched and fractured during the long ages of 

 elevation, and must have been e.xposed to the usual 

 destructive forces, which may even have kept pace with 

 the elevation during its late stages when differences of 

 level became pronounced. 



The coarse-textured granitoid gneiss making up the 

 batholith must have cooled at great depths and exceed- 

 ingly slowly. 



The. Raising of the Domes. 



Some curious dynamical problems are involved in tlie 

 raising of the domed mountains. It is conceivable that 

 fluid lava could be forced by the unequal pressure of 

 shifting mountain blocks through a suitable system of 

 pipes into cisterns, so as to form laccolithic domes, but 

 no such mechanism seems possible with batholiths. The 

 granite of the batholiths was plastic rather than fluid, as 

 shown by its having been dragged into the gneissoid 

 structure. The areas affpcted covered sometimes 1000 

 square miles. We know of no system of dykes to sprvo 

 as pipes or passages, of no solid floor beneath, of no 

 faulted blocks to provide the pressure. It is generally 

 assumed that the protaxial granites and gneisses in great 

 mountain ranges have risen because of the relief from 

 pressure beneath anticlines due to lateral thrust. It is 

 doubtful if these irregularly scattered ovals, sometimes 

 thirty miles across, can be adjusted to any system of 

 anticlines. 



Some years ago I ventured another explanation. 

 Granite is specificallv lighter than most of the green- 

 stones and schists of the Keewatin, and molten granite, 

 even if not at a very high temperature, is lighter than 

 the relatively cold rocks above it. If the rocks above 

 were unequally thick, so that some areas were less 

 burdened than others, it is conceivable that these differ- 

 ences in gravity might cause the granite to creep slowly 

 up beneath the parts with the lightest loads, while the 



NO. 2133, VOL. 84] 



overlying rocks sagged into synclines in the heavily loaded 

 parts.' 



Whatever their cause, these oval batholiths enclosed by 

 meshes of schist are the most constant feature of the 

 Canadian Archjean, though in many places erosion has 

 cut so deeply that the meshes have all but disappeared, 

 leaving only straight or curving bands of hornblende schist 

 enclosed in the Laurentian gneiss. Very similar batho- 

 lithic relations of the Laurentian with the Grenville series 

 of Eastern Ontario are described by Drs. Adams and 

 Barlow, tiiougli the batholiths are generally much smaller. 

 Batholithic mountains were typical of the Archaean in 

 North America, and, at least in some cases, also of 

 .Xrchjean regions in other parts of the world. 



Subdivisions of the Canadian Pre-Camhrian. 



Until recently the rocks of the Canadian Shield were 

 usually divided into three parts — the Laurentian, the 

 Huronian, and the .\nimikie and Keweenawan, the last 

 two being onh' doubtfully included in the pre-Cambrian. 

 These three divisions are still the only ones shown on the 

 latest general map prepared by the Geological Survey. 

 Lawson 's separation of the Keewatin as a lower group 

 than the Huronian was generally recognised as valid, 

 but in practice the subdivision of the two in mapping was 

 difficult, and was only carried out in detailed surveys. 

 His proof that the Laurentian was eruptive and later than 

 the Keewatin was accepted. 



As the classification adopted by the American geologists 

 in the Lake Superior region differed from that used in 

 Canada, a Correlation Committee w'as appointed five or 

 six years ago to draft a compromise, which runs as 

 follows :— 



Keweenawan 



Unconformity 



Huronian-! Middle 



f Upper (.^nimikie) 

 I Unconformity 



j Middle 



L'ncon 

 I Lower 



formity 



Unconformity 

 Keewatin 



Eruptive Contact 

 Laurentian 



This compromise system is now generally in use in 

 Canada, though if Canadian relationships alone were con- 

 sidered the Animikie would be separated from the 

 Huronian and placed closer to the Keweenawan, and the 

 Laurentian would be treated as consisting of eruptive 

 rocks frequently later in age than the Lower Huronian. 



The most natural classification for Canada would be 

 as follows : — 



Keweenawan 



Unconformity 

 Animikie 



Great Unconformity 



Huroniar|!r'PP^'' 

 ) flower 



Great Unconformity 

 Keewatin 

 Laurentian = post-Keewatin or post-Huronian granite 



and gneiss. 

 The laccolithic domes described on previous pages were 

 formed partly in the interval between the Keewatin and 

 the Lower Huronian. but mostly later than the howet 

 Huronian. Over much of the shield, however, our know- 

 ledge of the relations is not sufficient to separate the 

 mountain structures of the two ages. 



Let us now consider the history of the region during 

 the successive periods suggested above. 



Conditions during the Keewatin. 

 One naturally asks what the conditions were_ in Kee- 

 watin times before the earliest known laccolithic moun- 

 tains were raised. The granitic texture of the eruptives 

 implies ^ery slow cooling under great pressure. The old 

 interpretation of these rocks, following the usual concep- 

 tion of the nebular hypothesis, made them parts of the 



1 Bull, Ccv!. Soc 



, vol. 



