FBBBUAKr 7, 1908] 



SCIENCE 



227 



them, they are not convincing. Thus in the 

 table of data on page 237 relating to oviposi- 

 tion, the maximum, minimum and average for 

 any given species may be made up of more or 

 less complete observations on two beetles or 

 on a thousand; at all events, new observations 

 will change the figures. Data on the oviposi- 

 tion of L. decem-lineata which Mr. A. A. 

 Girault is about to publish will change the 

 aspect of this table very materially. 



It goes without saying that there is much 

 excellent material in Professor's Tower's work. 

 The observations on habits are most interest- 

 ing. A point well worth the attention of ex- 

 perimental biologists is that tropical species, 

 being less subject to fluctuating conditions 

 than those of more northerly regions, respond 

 more readily to change of environment. 



The work, along with other Carnegie pub- 

 lications, suffers very materially through the 

 absence of an index. 



Frederick Knab 



Washington, D. C. 



SPECIAL ARTICLES 



AGE OP A COOLING GLOBE IN WHICH THE INITIAL 



TEMPERATURE INCREASES DIRECTLY AS THE 



DISTANCE FROM THE SURFACE 



Kelvin's famous and epoch-making paper 

 on the secular cooling of the earth was pub- 

 lished in 1862.^ His problem was to find the 

 time which would elapse before a globe com- 

 pletely solid from center to surface and hav- 

 ing throughout a certain uniform initial tem- 

 perature would cool so far as to reduce the 

 surface gradient of temperature to any given 

 value. He assumed an initial temperature 

 of 3,900° C, a diffusivity of 0.011Y8 in c.g.s. 

 units and a final surface gradient of 1° C. in 

 2Y.'76 m. or 1° F. in 50.6 feet. These data 

 discussed by one of Fourier's theorems give 

 for the age of the earth 98 X 10° years. Kel- 

 vin, however, expressly directed attention to 

 the fact that the effect of temperature in 

 modifying diffusivities is almost unknown, 

 and that the original distribution of tempera- 

 ture is uncertain. He also referred to the 



' Trans. B. S. Edinburgh, reprinted in Thomson 

 & Tait, " Natural Philosophy," Pt. II., p. 468. 



great differences in the surface gradient of 

 temperature, which varies with the locality, 

 as he stated, from 1° F. in 15 feet to 1° F. in 

 110 feet. He, therefore, allowed very wide 

 limits in his estimate and placed the age be- 

 tween 20 million and 400 million years. 



In 1893 Clarence King made a very impor- 

 tant contribution to the subject'' by intro- 

 ducing the criterion of tidal stability. Mr. 

 Barus determined for him the melting point 

 of diabase in terms of depth. If in any hy- 

 pothetical earth consisting solely of diabase 

 the temperature in any couche were to exceed 

 the melting point of diabase, then tidal insta- 

 bility would set in, the crust would break 

 down and chaos would reign for the time 

 being. In a real earth the same result would 

 follow provided the couche were in a region 

 where diabase or equally fusible rocks are to 

 be expected. Excluding such cases. King- 

 found that the age of the earth could not 

 exceed 24 million years when Kelvin's values 

 for diffusivity and surface gradient are as- 

 sumed. He also found that the correspond- 

 ing initial temperature of such a globe would 

 be 1,950° C. 



Kelvin's last paper on a cooling earth' was 

 read in 1897 and he there stated that after 

 having worked out the problem of conduction 

 of heat outwards from the earth by an elab- 

 orate method, he was not led to differ much 

 from Clarence Kings's estimate. This he 

 adopted as the most probable age and reduced 

 his limits to between 20X10° and 40X10° 

 years. 



While King's earth is tidally stable, I con- 

 fess that his solution of the problem seems to 

 me to be fatally defective. He himself gives 

 a temperature curve for the same earth at an 

 age of 15 million years and this earth shows 

 a couche at a temperature above the melting 

 point of diabase, this layer extending from a 

 depth of 34 miles below the surface to 66 

 miles. According to Laplace's law of densi- 

 ties these two levels correspond respectively 

 to densities of 2.85 and 2.93, and it seems 

 certain that the material must consist chiefly 

 of basaltic rocks. Thus the 15-million-year 



"Am. Jour. Sci., Vol. 45, 1893, p. 1. 



'Tram. Victoria Institute, Vol. 31, 1899, p. 11.- 



