FORM OF THE CONTINENTS— WATTS 193 



as sediment upon it the marine stratified rocks which are our chief 

 witness of the rhj^thmic advances of the sea. 



This condition, however, cannot be permanent, for by convection 

 of the fluid basic substratum, supplemented by the influence of tides 

 within it, and tlie slow westward tidal drag of the continental masses 

 toward and over what had been ocean floor, there will now be dissipa- 

 tion of its heat, mainly into the ocean waters, at a rate much faster 

 than it has been or could be accumulated. Resolidification ensues, and 

 again there are two main consequences. First, the stratum embedding 

 their roots having now become more dense, the continental masses 

 rise, and as they do so the ocean waters retreat from their margins 

 and epicontinental seas, leaving bare as new land, made of the recently 

 deposited sediments, the areas previously drowned. Secondly, the 

 expanded crust, left insufficiently supported by the withdrawal of 

 shrunken substratum, will suffer from severe tangential stress, and, on 

 yielding, will wrinkle like the skin of a withering apple. The wrinkles 

 will be mountain ranges, formed along lines of weakness such as those 

 at continental margins; and they will be piled up and elevated to 

 suffer from the intense erosion due to water action upon their exposed 

 and upraised rocks. 



In this, again, we have a mechanism which suj^plies what was 

 needed by Suess, and one, moreover, which secures the required rela- 

 tionship between continental and mountain movement, between the 

 broader extensions of continental land and the growth of mountains 

 with their volcanoes and earthquakes and the other concomitants of 

 lateral thrust. 



Thus a thermal cycle may run its full course from the solid sub- 

 stratum, through a period of liquefaction accompanied by crustal 

 tension, back to solidification and an era of lateral stress; and the 

 stage is set for a new cycle. 



Prof. Arthur Holmes, in checking Joly's calculations, has concluded 

 that the length of the cycles in a basic rock substratum should occupy 

 from 25 to 40 million years, a period much too short to fit the major 

 periods of mountain movement, as determined by him from the radio- 

 activity of minerals contained in the rocks. On this evidence the 

 Alpine movement should date back from 20 to 60 millions of years 

 ago, the Hercynian 200 to 250 millions, and the Caledonian from 350 

 to 375 million years. 



In a preliminary attempt to modify Joly's hypothesis Holmes 

 postulated the occurrence of similar, but longer cycles (magmatic 

 cycles) in a denser, ultrabasic layer underlying the basic one, tlie 

 rhythm of which would be nearer to 150 million years. The shorter 

 cycles due to the basic layer are held in part responsible for periods 

 of minor disturbance, and also to account for the individual varia- 



