Mr. Powell on Li^ht and Heat, 215 



effect above described not to have been sufficiently excluded. The 

 larger size of the bulb may have partly caused the effect to have 

 been so much less with this than with the smaller instrument, the 

 diameters of the bulbs being respectively 0.35 and 0.6 inch, be- 

 sides the additional causes in operation on the small instrument 

 tending to increase the effect before noticed. 



(4.) The peculiar effects above described on the larger instru- 

 ment, appear to be due to some greater action on the plain bulb ; 

 in proportion as this interfering cause is excluded, the light from 

 incandescent metal is capable of displaying its heating effect, and 

 the action on the plain bulb cannot be ascribed to light, but must 

 depend upon some peculiar action of simple radiant heat. It is 

 evident that if there were any thing tending to make the simple 

 heat act more on the plain bulb, its eflfects would be displayed 

 when the screen was removed. This might be the case from the 

 circumstance that the plain bulb was of rather smaller diameter, 

 and the glass probably thinner than the black enamel. Tt might 

 also be a question, whether a greater apparent effect on the plain 

 bulb might not be in part occasioned by the greater expansion of 

 the black glass. 



Thus, upon the whole, we may regard these apparent anomalies 

 as tending to establish still more firmly the essential distinction 

 between these two heating agents at first pointed out. 



(5.) To those who have been led to attend accurately to the na- 

 ture of the instrument in question, it will be needless to make any 

 further observations on the several causes of inaccuracy affecting 

 its indications, which have been recently pointed out by several 

 experimenters. The use here made of it, however, is not of such 

 a kind that the results would be materially influenced by any such 

 inaccuracy, and the main facts have been established upon inves- 

 tigations conducted in a very different manner. 

 ' (o.) The following are a few of the experimental results above 

 referred to :-^ 



