264 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1937 



have doubted the validity of this conclusion. One, for example, 

 prefers to explain the high central density by the combination of 

 atoms under the conditions of high pressure and temperature to form 

 heavier atoms of various kinds. Others have inclined to the belief 

 that under sufficiently high pressures the structure of all solids will 

 collapse, leaving a material of the same chemical composition and 

 greatly mcreased density. 



This reminds one of the heated discussion carried on some years ago 

 in the columns of a magazine devoted to popular science, as to whether 

 the water at the bottom of the ocean was as dense as cast iron. One 

 faction contended that engineering data showed that materials or 

 structures subjected to sufficient compression in a testing machine 

 invariably failed by crushing, and that at considerable depths in the 

 ocean the water would crush, or cave in upon itself, and thus become 

 as dense as a heavy metal. The principle that was overlooked in this 

 contention was that it is relatively very difficult for pure hydrostatic 

 pressure to make a structure collapse. It is true that we must be on 

 guard against applying the experience and conclusions pertaining to 

 a limited range of pressures and temperatures to the extreme con- 

 ditions prevaiUng in the interior of the earth. 



On the other hand, various considerations indicate that the postu- 

 lated transmutation of elements and collapse of crystal structure will 

 take place only under pressures and temperatures of a higher order 

 of magnitude than those existing in the earth, that is to say, in the 

 interior of stars rather than planets. 



Passing over the intermediate zone for the moment, let us turn our 

 attention to the outer layer, commonly called the crust. This term 

 dates back to the time when the earth was thought to consist of a 

 thin solid crust surrounding a molten interior. But the notion of a 

 true crust floating on a thm Uquid is now abandoned, although the 

 word is still used to designate the outer layer, perhaps 40 or 50 kilo- 

 meters thick, with properties very different from those of the material 

 below it. 



From geologic studies it has long been knouTi that the accessible 

 part of the crust consists largely of granite. When the measurements 

 of the compressibilities of rocks, already referred to, were made and 

 the velocity of longitudinal vibrations in typical granites found to be 

 5.6 kilometers per second, it was therefore a source of gratification 

 to find that this was precisely the speed found by the seismologist 

 for the longitudinal waves in the outermost parts of the crust. From 

 seismologic data we know that tliis granitic layer varies in thickness 

 from place to place. In continental areas it may be as Uttle as 10 

 kilometers or as much as 30 kilometers in thickness, while in the great 

 ocean basins it appears to be entirely missing. 



