of the Glacial Epoch. 333 



applicable to the determination :: the past influence of internal 

 heal upon the surface temperature of the ocean. There are four 

 circumstances which require consideration in attempting to 

 realize the thermal condition of the ocean during the cooling of 

 the earth's crust. These are, first, the conduct:".::-- ;: water; 

 secondly, its convection ; thirdly, its power of penetrating into 

 any cliasrns or dislocations that might be formed from time to 

 time in the floor of the ocean : and fourthly, its specinc heat. 



"With regard to the conductivity of water, the determinations 

 of Desprets* show that ; compare:: with the conductivity of the 

 solid crust :: the earth, it may be regarded as a vanishing quan- 

 :.:; But this almost total a: c :: :: :: ::::.::: ting -power is more 

 than compensated for by fluid convection. If the ocean were 

 free from lateral currents, however even this convection would 

 not be much superior to the conductivity of granite in trans- 

 mitting he:: :: the surface ; ::: :: we assume the average 

 of the sea to be nve miles .:. even if its temperature at the 

 bottom were 100 : C. and that at the surface 15° C. ire should 

 only have a difference of '0032 : C. for each foot of ascent. 

 There are no determinations :: the velocity :: : : ::vection in 

 water for given differences :: temperature, but, for the small 

 foot-difference contemplated, it must be almost inconceivably 

 small. Nevertheless it can scarcely be doubted that the j lai 

 and equatorial currents aid this vertical convection to such 



: as to render the total :: ansfei :: heat :: m the floor to the 

 surface of the ocean considerably more rapid than that which 

 would take place through a solid stratum of granite of the 

 thickness. I have endeavoured roughly : :::mine the rela- 

 tive powers of water and granite in thus transmitting heat. 1 nt 

 this purpose a 6-inch cube was cu .: :: [rranite, and 



a vessel of tin plate we- red capable of holding a sin 



cube of water c . :th a layer of oil ^gth of an inch thick I 

 prevent evaporation. B ith cubes were place:! upon an iron 



ibelow by an atmosphe: : : ate ::. tl vertical sides being 

 : ted from loss of h ral folds f flannel. Late 



. which wei to . the bulb 



of a thermometer was imbedded, and the time required to elc 

 the surface temperature 10 : C. was noted. The following are 

 the results of these experiments : — 



lime required to raise temper:.:- »f surface"! _* ™ 



of granite cube through 10 : C. J ~~ 



Time required to raise temperature of surface "\ 

 of water cube through 10"- C. . , . J* ~~ 



* Aam. ie Chim. et de\Thys. vol. Lsxi. p. 206. 



