of the radiant heating Effects from terrestrial Sources. 443 



(15.) In these experiments it will be evident upon inspection 

 that the ratio of the effects produced on the white and black 

 bulbs is in every instance considerably greater, when they 

 were affected only by that part of the total heating influence 

 which is transmitted through a transparent screen, than when 

 they were exposed to the whole. This then would indicate, 

 that on the removal of the screen some new heating power was 

 brought into action, which affected the ratio by the addition 

 to each of its terms, of quantities in a ratio expressed by that 

 of the difference of the exposed and screened results above 

 given. This ratio is evidently one differing a little from equa- 

 lity, and agreeing nearly with that of the diameters of the bulbs 

 inversely. 



(16.) The experiments now detailed will probably be con- 

 sidered sufficient to substantiate the conclusion; but in re- 

 searches of this kind, where great numerical precision is unat- 

 tainable, it seemed desirable to give the experiments that con- 

 firmation which they wanted in point of intrinsic accur-acj', by 

 frequent repetition and variation. With this view I made a 

 great number of trials with a large differential thermometer : 

 the bulbs were about one inch in diameter, and nearly three 

 inches apart. The bore of the tube was about yoth of an 

 inch. Many of the experiments made with this instrument I 

 shall not mention, as, although all agreeing to confirm my 

 former conclusions, they were complicated by several unne- 

 cessary conditions. 



(17.) In order to obtain results in the most simple manner, 

 it was desirable to get rid of any action on one of the bulbs, 

 and to expose only the other; the instrument thus acting sim- 



Ely as an air-thermometer. The effects on each bulb, one 

 eing painted with Indian ink, and the other coated with white 

 silk pasted on, when exposed, might thus be compared with 

 those through a glass screen. I first tried the experiment by 

 placing the bulb in the focus of a spherical tin reflector about 

 six inches diameter: by this means the source of heat could 

 be placed at a sufficient distance to preclude any effect from 

 the glass screen. 



. (18.) The experiment was again varied by placing a large 

 opake screen before the instrument, in which was an aperture 

 througii which one bulb might be exposed. To this aperture a 

 piece of glass could be applied. Each bulb was presented both 

 with and without the glass. 



(19.) In all these experiments it is evident, that any heating 



effect arising from the screen would tend to diminish the ratio 



of the black and white effects ; and this not being allowed for 



in the statement of the result, the difference between this ratio 



3 K 2 and 



