262 PKOFESSOR JAMES GETKIE, LL.D., D.C.L., F.B.S., ETC., ON 



interest, and yet with disappointment at seeing so little new light 

 thrown on the difficult and interesting subject of secular changes 

 of climate. 



Before speaking of the general question, there is a special point, 

 on which, though not myself a geologist, I must venture to differ 

 from Professor Geikie. He says : — " No one acquainted with the 

 physical features and geological structure of Scotland and Norway 

 can doubt that the valleys which terminate in fiords are of great 

 geological antiquity. Their excavation by fluviatile action certainly 

 dates back to a period long anterior to the Ice Age." On general 

 grounds I think this statement is partly misleading. Not very 

 many sea-coasts are cat up into fiords ; and it cannot be a mere 

 coincidence that fiords have been formed chiefly on those coasts 

 where glaciation is most favoured by the geographical conditions, 

 namely, on mountainous coasts, in high latitudes, and where ex- 

 posure to prevailing west winds from the ocean promotes an 

 abundant snowfall. Norway, Scotland, and the west of Ireland, 

 presents such coasts ; but the most conspicuous instances will be 

 seen by a glance at a map of the world, to be at the northern and 

 southern ends of the western coast of the American continent. 

 From Vancouver's Island northward, and from Chiloe southward, 

 the coasts of the continent are cut up into fiords and islands by 

 sounds which are submerged valleys ; while in the lower latitudes, 

 both northern and southern, the coast, from Vancouver's Island to 

 Chiloe, is remarkably unbroken. 



The connection between glaciation and the formation of fiords is 

 obvious enough. Most valleys have been excavated ; and these, 

 except some which have been eroded by the sea, are due either to 

 fluviatile or to glacial action. Mountain valleys excavated by running 

 water are in general deep and narrow — the most conspicuous 

 instances are the canyons of the Colorado, and the Via Mala in 

 Switzerland, which is a canyon — and, though on a much smaller 

 scale, the ravine-like valleys of the so-called Saxon- Switzerland are 

 of this class. Mountain valleys excavated by glaciers are on the 

 contrary deep and wide ; and it appears to be generally agreed that 

 most of the valleys of our European mountains are of this origin. 

 When such a valley descends into the sea it becomes a fiord. It 

 may be the fact that most of the greater valleys of Norway and 

 Scotland existed as river valleys before the Glacial Period, but if so, 

 during that period they became filled with glaciers, which, by their 

 excavating action, gave the valleys their present form and contour. 



I am fully convinced that no merely geographical changes can 

 possibly account for the glacial climate ; and I agree with Mr. Croll 

 that its causes were astronomical. But I think he has failed to 

 explain rightly how these causes operated. 



I must here point out that the extent of glaciation depends in no 

 degree on mean temperature, but exclusively on summer tempera- 

 ture. The " snow-line " is the line of summer snow, and theory and 



