464 Serviss — Internal Temperature Gradient of Metals. 



radiated beat at the constant rate of about 100 gram calories 

 per hour per gram. Applying this, Strutt's most active material 

 would radiate 9x10"" gram calories per second per gram. 

 Substituting this aud the other constants for iron, we have, 

 with o = 5 cm , 



15XlO s t = 8025 — r, 



from which we get, 



r 



15* 



t 



cm 



8-025 X10- 5 



5-35X10 



1-0 



8-024 





1-4 



8-023 





2-1 



8-021 





4-4 



8-006 





5-0 



8-000 



5-33 



The junctions of -my thermopile came at the third and the 

 fifth of these points, and it will be seen that the difference 

 between these could certainly have been detected with niy 

 apparatus. Dr. Thwing gives the temperature of iron at r = 2*1 

 as 3-2X10" 5 , i. e. his iron was about six times as radio-active as 

 Strutt's most active substance, if we grant that the outer junc- 

 tions of the thermopile in the air bath were at zero, and that 

 the entire leak in Strutt's electrometer was due to the radio- 

 activity of the specimens tested and not partially due to other 

 causes, as the emanations from the earth. 



It is interesting to compare, in this connection, the gradients 

 in the same sphere of iron, heated to 20° throughout and 

 allowed to cool in the air bath at 0° ;* we set 



r 



5 tnin. 



li -III IjilC 



15 rnia. 



1 hour 



2 hours 



3 hours 



6 hours 







19-°221 



17-687 



12-450 



o-ioio 



0-00717 



2-57x10" 



1-4 





17-683 











2-1 





17-679 











4-4 





17-645 











5-0 



19-161 



17-632 



12-410 



0-1007 



0-00715 



2-56x10 



)iff. 



0-060 



0-055 



0-040 



0-0003 



000002(2) 



0-08x10' 



* 



0-31 



0-31 



0-32 



0-32 



0-31 



0-31 



Similar computations have not been made for lead, because, 

 so far as I can find, experiments have failed to measure the 

 radio-activity of lead. 



These tables show the decided discontinuity at the surface ; in 

 fact, in both cases, only about 0'3 of 1 per cent of the total differ- 

 ence of temperature between the center and the air bath oc- 

 curs in the sphere itself ; and moreover, this fraction remains 

 very nearly constant throughout the range computed. The 

 poorer the surface conductivity, other things remaining equal, 



*W. E. Byerly: Fourier's Series, etc., pp. 116-122. Bostou, 1893. 



