510 Mr. J. J. Waterston on Solar Radiation. 



A represents the logarithm of 2 ; so that, c being a constant, 



we have log 2 = — . In the curve that represents the cooling 



in air, we may assume a small arc of it to coincide with a true 

 logarithmic curve, or the curve of cooling in a vacuum, and we 

 have to find the value of r. the ordinate of the true logarithmic 

 curve at the given point. 



The logarithmic curve is defined by the equation 



clog-i=f 1 -^ = 1 -- y log- 1 . 



loa:2 





Let 



, =(r 1 -0°l). 



To find the value of r v corresponding to a given r a (r v vacuum, 

 r a air), we require to compute the value of t l — t Q , employing 

 A a = 294— g Vr in the equation 

 A„ , r 



lo §7^ ^TT=^i-^o; 



log2 x °(r a -0°'l)" tl l ° 



then, with this value of t 1 — t , and with A =294, find the value 

 of r v in the equation 



log 2 c (r L ,-0°-l) 

 The direct equation is 

 A a {logr a -log(r a -0 o -l)}=A o {log7^-log(r„-0 3 'l)} 

 and 



— 1— ?T7VT v r„=l 



A D 294 v/ "~ x 17-131 



Hence r v may be ascertained by inspecting the differences of 

 a table of logarithms; and it was from these that the Scale, 

 fig. 4. Plate V., was constructed for reducing the values of r 

 taken in air to what they would be if taken in a vacuum, where 

 the emission of heat was by radiation alone. 



The cooling of the sun-thermometer in air when fixed in its 

 place in the tube, as in fig. 1, was found to be exactly the same 

 as when fixed in the cylinder, fig. 2, unexhausted. 



A chemical thermometer with cylindrical reservoir was tried 

 in the vacuum-bath^, and the cooling was found to take place 

 exactly in the logarithmic curve. It is difficult to adjust the 

 vacuum-bath in time to observe a high value of ?•, but good 

 observations were obtained from r=190° downwards; so there 

 is little doubt that the law of cooling by radiation is general and 

 independent of the shape of the cooling body. I purpose extend- 



