442 Mr. Wheeler on the Application of Photometry 



The results in this last column differ so widely from Mr. 

 Potter's, that I was led to examine several times over the 

 whole calculation ; but have been able to detect nothing which 

 can have led to error. I can therefore merely give my re- 

 sults as they stand, and those interested in the subject may 

 possibly be able to throw some light upon the point by further 

 examination. 



I will only add, that with regard to the undulatory theory, 

 Mr. Potter's calculation of the ratio of illumination from that 

 theory, viz. 1*1538, approaches so much more nearly to agree- 

 ment with my results than he found it to do with his, that 

 perhaps we shall find the error (as there must be one some- 

 where) will in time show that there is in reality a still more 

 close accordance between theory and experiment. Indeed, 

 I will add that as Mr. Potter's calculation is founded upon 

 his estimate of ^th of the incident light being reflected from 

 glass, so from some trials of my own, (which I do not doubt 

 are of far inferior accuracy,) I found the proportion more 

 nearly g-'^th. Now this was quite independent of the last ex- 

 periments. And it is curious, that on calculation with this 

 number, in the way adopted by Mr. Potter, I find the re- 

 sulting ratio from theory about 1-25, which accords exactly 

 with these experiments. 



I made other sets of experiments, as Mr. Potter did, to 

 compare the illumination from reflection at the surfaces of se- 

 veral different substances, with the view of comparing the re- 

 sults with the expression for such illumination derived from 

 the undulatory theory. 



In the first instance I compared the effects of crown glass 

 and of a diamond possessing a tolerably good plane surface, 

 to the area of which that of the glass was carefully equalized : 

 we have then to find if (for glass), as before, by Mr. Potter's 

 formula, and since we here equalize the lights, or make 



I = I' we have for diamond i m i' --. 



cos<$ 



