G. K. Gilbert — Inculcation of Scientific Method. 295 



sarj to suppose that the sea floor locally sank down as rapidly 

 as the sediments were added. Conversely there is reason to 

 believe that the adjacent continent, which by erosion furnished 

 the sediment, rose up as rapidly as its surface was degraded. It 

 is a favorite theory — at least with that large division of geolo- 

 gists who consider the interior of the earth as mobile — thnt the 

 sea-bottom sinks in such cases because of the load of sediment 

 that is added and that the land is forced up hydrostatically be- 

 cause it is unloaded by erosion. A similar theory might ex- 

 plain the up-arching of the desiccated bed of Lake Bonneville, 

 for the unloading of 1000 feet of water from an area more 

 than one hundred miles across would give to the supposed 

 liquid interior an irresistible uplifting force. This was the first 

 explanation to suggest itself. 



The second suggestion did not spring from any geological 

 theory consciously retained in memory, but I have since sus- 

 pected that the germ of the idea may have been caught from a 

 passage in Croll's 'Climate and Time.' It is this: The geoid 

 of which the ocean's surface is a visible portion is not an ellip- 

 soid of revolution, but differs from that symmetric surface by 

 undulations which depend on local inequalities in the density 

 and in the superficial configuration of the earth. The water 

 level is everywhere normal to the plumb-line, but the plumb- 

 line, as geodesy has shown, is subject to local deflection. Now 

 the ocean itself is one of the attracting factors, and if the 

 ocean were to be removed, the geoid would thereby be modi- 

 fied. The surface of Lake Bonneville was part of a geoid at a 

 higher plane than that of the ocean surface, and the removal of 

 the water of the lake unquestionably modified the local form 

 of the geoid. Only at first blush the cause seems too small for 

 the effect observed. 



The third suggestion relates to the distribution of tempera- 

 tures beneath the surface of the earth. It is well established 

 that the inner parts of the earth are extremely hot. The outer 

 surface is relatively cool, and in the intermediate region there 

 is a gradation of temperatures. The isogeotherms, or planes 

 of equal temperature, are not even surfaces, but undulate in 

 response to variations of conductivity and of superficial tem- 

 perature. At the poles, where the external surface of the crust 

 is exceptionally cold, the isogeotherms lie lower down than in 

 warmer latitudes ; and if a portion of the earth's surface un- 

 dergoes a permanent change in temperature, the influence of 

 this change is propagated slowly downward through the crust, 

 and the isogeotherms are locally raised or lowered. Where they 

 are raised, the crust is locally expanded, and its surface is up- 

 lifted; where they are depressed the surface of the crust sub- 

 sides. If, therefore, it can be shown that the temperature at the 



