68 SECTIONAL ADDRESSES. 



invaded by granite batholiths ' ; and again : ' The major part of the folding 

 and faulting of the Palaeozoic and late Precambrian strata took place 

 during the Devonian interval of orogenic disturbances.' 



There is, it must be admitted, a minor, perhaps only an apparent, 

 delay in the Caledonian history of Canada as compared with that of 

 Britain. In Canada certain important marine limestones that are 

 involved in the mountain folding are, according to present-day termino- 

 logy, referred to the Lower Devonian ; while the Gaspe Sandstone, that 

 seems to play a role comparable with that of our Lower Old Red Sand- 

 stone, is generally spoken of as Middle Devonian. Perhaps, as already 

 said, this lack of harmony may be only apparent, for the Gaspe Sandstone 

 agrees closely in its flora and fish fauna (if we include the Campbellton 

 fishes) with our own Lower Old Red Sandstone. Indeed, not many years 

 ago, the Gaspe Sandstone was treated by Williams and others as Lower 

 Devonian. It was transferred from Lower to Middle by Clarke and Kayser 

 on the basis of comparisons between the marine successions of America 

 and Rhineland. Possibly we shall some day regain the correlation of the 

 Gaspe Sandstone with our ' Lower ' Old Red Sandstone by accepting 

 Barrois' transference of our Downtonian from Silurian to Devonian. 



Let us now go back for a moment to 1843, when Logan started the 

 Geological Survey of Canada. By this time Hall and his colleagues had 

 already determined the main stratigraphical features of the flat-ljdng 

 Palaeozoic rocks of Laurentia as exposed in the western portion of New 

 York State. The succession there starts with Potsdam Sandstone of 

 Upper Cambrian date, and continues upwards through a long quasi-con- 

 formable sequence into the Carboniferous of Pennsylvania. Logan had 

 no difficulty in applying Hall's classification to the rocks of the St. Lawrence 

 Lowlands ; but at the south-eastern margin of these lowlands, along the 

 course of the Champlain and St. Lawrence, he saw the familiar Ordovician 

 of Laurentia passing beneath folded mountain-rocks, that at first seemed 

 unidentifiable, whether on the score of lithology or fossils. Faced with 

 this difficulty he was, for several years, content to date the mountain 

 formations by the law of superposition. So long as they yielded only a 

 few scattered fossils, this seemed quite reasonable. Barrande, looking 

 from across the Atlantic, might claim an occasional trilobite as of Cambrian 

 date ; but, naturally, local observers could not understand the sanctity 

 that Barrande attached to trilobite successions, remembering the theory 

 of ' colonies ' which he himself had introduced to account for graptolite 

 recurrences. In 1860, however, fossils were rediscovered in abundance 

 in the Levis exposures that overlie the top of the Ordovician in the 

 neighbourhood of Quebec. Many of these fossils were of Cambrian, 

 others of early Ordovician types. The number of forms was so great 

 that to apply the theory of ' colonies ' to account for their position would 

 have been tantamount to throwing to the winds all faith in palaeontological 

 stratigraphy. Accordingly Billings, the Palaeontologist of the Canadian 

 Geological Survey, transferred the Levis rocks to a low position in the 

 Ordovician, where they remain to this day. 



Billings was working in close touch with Logan, who thoroughly 

 appreciated the significance of this stratigraphical revolution. On 

 December 31. 1860, Logan addressed a long letter to Barrande, and told 



