88 Mr. Powell on Solar Light and Heat. [Aug. 



The former set of experiments gives the ratio in the focus 

 somewhat less than in the diffuse rays, the latter somewhat 

 greater. "We may, therefore, fairly infer, that the ratios are very 



nearly equal in the two cases. The means are ^jg and ^^. 



The small difference between these two sets of experiments 

 must be attributed to the impossibility of laying on the coatings 

 in the second instance so as to be sure that they are of precisely 

 the same thickness, roughness, &c. as in the first ; but the dis- 

 agreement is so small as to show that such an equality was as 

 nearly attained as perhaps was possible. 



(37.) Being in possession of these two sets of experiments, 

 we might have deduced the result without any reference to the 

 formula. Proceeding by this method, therefore, we may ascer- 

 tain the accuracy of the data, and thus also tend to show how 

 far the other investigations here made are to be relied on. It 

 will be evident that we have in these two sets, with the foci, 



it i e <P*iV ' 



Hence, as before, g-^ = ^, 

 And .\~r = tH nearly, 



Pt k 2-4 J 



results which agree very closely with those otherwise obtained. 



(38.) In order to separate from this result the value of-, I 



ascertained j by independent experiment : heating the two 



thermometers to the same point, and then observing theirrates 

 of cooling, as follows : 



