16 



Rev. Mr. PoioeWs Remarks on some of [July, 



Here the apparent rate of cooling when the bulb was plain 

 was at least equal to that when coated, or perhaps even greater. 



This result agrees with the last, and in both cases it is evident 

 that if no other cause interfered, we ought to find the coated 

 bulb cool considerably faster than the plain ; but if no difference 

 appear, or a contrary action take place, we must ascribe to the 

 expansion of the glass an action equal to or greater than that of 

 the opposite kind due to the radiating power of the coating. 



Mr. Ritchie, in the same paper, has maintained his position 

 by various other experiments. He used as screens pieces of a 

 glass bulb blown to a great degree of tenuity, and compared the 

 effect of such a screen when transparent with that exhibited 

 when it was opaque ; care being taken that the nature of the 

 surface was the same in each case, by forming a compound 

 screen of several thicknesses of glass, the opaque coating or 

 lamina being interposed. But in all these experiments the effect 

 was observed by means of a large differential thermometer; and 

 considering the acknowledged uncertainty which attaches to 

 the action of such an instrument, we can hardly feel implicit 

 confidence in any results deduced with it, unless corroborated 

 by those obtained by a more unexceptionable mode of observa- 

 tion. I have accordingly tried experiments of this kind, using 



