160 



SCIENCE 



[N. S. Vol. XLI. No. 1048 



be called the euteetic level because the 

 additional temperature necessary to fusion 

 would there be a minimum. The question 

 then arises what relation may be supposed 

 to exist between the euteetic level and the 

 level of compensation. 



In computing the temperature distribu- 

 tion of a cooling globe which owes a part 

 of its heat to compression, or to initial tem- 

 perature, and another part to radioactivity, 

 it is necessary to proceed by trial and error, 

 or to test various assumptions and consider 

 which best fits the facts. I have assumed 

 various ages and computed other condi- 

 tions corresponding to the actual heat 

 emission of the globe. These other eondi-. 

 tions are the depth of the euteetic level, 

 the thickness of the radioactive shell (sup- 

 posed uniform) and the proportion of the 

 surface gradient due to radioactivity. Two 

 cases are of special interest, the assumed 

 ages being 68 million years and 1,314 

 million years. 



For the lower age the euteetic level is at 

 a depth of 121 kilometers, and thus coin- 

 cides with Hayford's compensation level, 

 the radioactive layer is 2.58 kilometers thick 

 and radioactivity supplies % of the surface 

 gradient or of the earth's heat emission. 

 For an earth 1,314 million years old the 

 euteetic level lies at 300 kilometers, the 

 radioactive layer is 12 kilometers thick and 

 just Ys of the surface gradient is due to 

 radioactivity. In this ancient earth the 

 highest temperature excess due to radio- 

 activity would be found at and below the 

 bottom of the active layer and would 

 amount to only 106°. This is not much in 

 comparison with the temperature of lavas 

 and, if this age is the highest worth con- 

 sidering, most of the earth's heat must be 

 due to compression. 



So great an age as 1,314 million years 

 seems incompatible with other features of 

 the problem. This age implies that a thick 

 shell extending from the compensation level 



downward to and beyond the euteetic level, 

 a shell more than 200 kilometers in thick- 

 ness, has cooled after solidification through 

 an average temperature interval of about 

 600°. Now the geodesists have shown that 

 at the compensation level the strains must 

 be small, and I have given reason for be- 

 lieving these strains even smaller than those 

 computed by the geodesists. But I hold it 

 impossible that a layer of rock 200 kilo- 

 meters thick can cool 600° without setting 

 up large strains. 



On the other hand, no such difficulty 

 arises in the case of an earth 68 million 

 years old, for it is easy to show that only a 

 very small amount of cooling has occurred 

 below its euteetic level. Furthermore, in 

 this case the level of compensation acquires 

 a definite and intelligible physical inter- 

 pretation. Local fusion would bring about 

 compensation. Where, then, should we 

 look for compensation if not at the euteetic 

 level? 



In such speculations as this some lati- 

 tude must be allowed. If, as the geodesists 

 suspect it may be, the compensation level 

 is as deep as 140 kilometers, and if this is 

 also the euteetic level, the earth is 100 

 million years old, the radioactive layer is 

 4.74 kilometers thick and 26 per cent, of 

 the heat emitted by the earth is of radio- 

 active origin. 



It has often been asserted that the dis- 

 covery of radioactivity indefinitely pro- 

 longs the probable age of the earth. To me 

 it seems that the determination of the level 

 of compensation limits both the age of the 

 earth and the amount of radioactive matter 

 in its outer shell. Geoege F. Becker 



U. S. Geological Survey 



THE CONSTITUTION OF TEE ATOM^ 



The subject of the constitution of the atom 



has come into extreme prominence — great ad- 



1 From the address of the president of the Eoyal 



Society, Sir William Crookes, at the anniversary 



meeting on November 30, and printed in. Nature, 



