194 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1936 



tions in effect, duration, and intensity of the larger ones. Each of the 

 later movements has also evidently been limited and conditioned by 

 the results of foregoing ones, and especially by areas of fracture and 

 weakness on the one hand, and by large stable masses composed of 

 rocks intensely consolidated, or already closely packed, on the other. 

 More recently Holmes has developed the possibility that the loss 

 of heat is mainly due to convection in the liquid substrata, and that 

 convection is the leading cause of the drifting and other movements 

 of the crust, and the disturbances that have occurred in it. He says : 



Although the hypothesis iuvolving subcrustal convection currents cannot be 

 regarded as established, it is encouraging to find that it is consistent with a 

 wide range of geological and geophysical data. Moreover, it is by no means 

 independent of the best features of the other hypotheses. It requires the local 

 operation of thermal cycles within the crust, and it necessarily involves con- 

 traction in regions where crustal cooling takes place. It is sufficiently complex 

 to match the astonishing complexities of geological history, and sufficiently 

 startling to stimulate research in many directions. 



The phenomena are difficult to disentangle as the number of operat- 

 ing causes has been so great and many of them are not fully under- 

 stood. But, underlying them all there is unquestionably the pulse 

 within pulse which Suess saw and of which Joly pointed the way 

 to explanation. 



The view at which we have arrived is neither strictly uniformi- 

 tarian nor strictly catastrophic, but takes the best from each hypo- 

 thesis. As Lyell showed, most of the phenomena of geology can be 

 matched somewhere and sometime on the earth of today ; but it would 

 appear that they have varied in place, intensity, phase, and time. 

 And, as Lyell w^as driven to accept evolution to explain the history 

 of life on the earth, so must we employ the same word to express the 

 life processes of the earth itself, as was suggested by Huxley in 1869 

 and strongly advocated by Sollas in 1883. 



The contrast in outline and structure between the Atlantic and 

 Pacific Oceans had long been noted when Suess formulated and used 

 the differences as the basis of his classification. 



The Pacific is bounded everywhere by steep slopes, rising abruptly 

 from profound ocean depths to lofty lands crowned with mountain 

 ranges, parallel to its shores and surrounding its whole area. On 

 the American side the Coast Range is continued by the Andes. On the 

 Asiatic side chains of mountainous peninsulas and islands, separated 

 from the continent by shallow inland seas, extend in festoons from 

 Kamchatka and Japan to the East Indies, eastern Australia, and New 

 Zealand. This mountain ring, as Charles Lapworth said, "is ablaze 

 with volcanoes and creeping with earthquakes", testifying that it has 

 been recently formed and is still unfinished. 



The Atlantic Ocean, on the other hand, is not bordered with con- 

 tinuous ranges, but breaks across them all : the Scottish and Welsh 



