on the terrestrial Rays that occasion Heat. 517 



heat were occasioned by the same rays, would produce a stop- 

 page of 485 rays of heat ; but we find that our glass stops no 

 more than 362, so that 1 23 rays cannot be accounted for by 

 this hypothesis. To this we should add 66 invisible rays, (that 

 is, 394 x ,167,) which, according to our 160th experiment, this 

 glass also intercepts. But the 107th experiment, if we reject 

 the hypothesis, immediately explains the difficulty; for here 

 we plainly see, that only 7 1 rays of heat of the refrangibility 

 of red light are stopped, whatever may be the stoppage of that 

 light itself. 



A yellow glass stops 819 rays of light: these will occasion 

 a stoppage of 496 rays of heat ; but this glass intercepts only 

 333, and therefore 163 rays of heat must also remain unac- 

 counted for. And, turning to the 155th experiment, we find 

 that yg rays, or j- of the 394 allowed to be invisible ones, are 

 also to be added to that number. 



If in the results of our second table we have had an excess 

 of heat, which the last hypothesis would not account for, we 

 shall, on the contrary, meet with a considerable deficiency, 

 when we come to consider those of the third table. 



For instance, our tube filled with well-water, including the 

 glasses at the end, intercepted 211 rays of light. These, at the 

 rate of 606 to the thousand, would produce only a stoppage of 

 128 rays of heat; but here we find no less than 558 of them 

 intercepted. To evade the pressure of these consequences, it 

 may be said, " that as before every invisible ray was supposed to 

 ,$ have been transmitted through glasses, so they may now be all 

 " intercepted by liquids/' And, granting this also to be possible, 

 though by no means probable, for the great extent of these 

 researches has not allowed sufficient time for many experiments 

 mdccc 3 X 



