2U Dr. H. R. Wright on the Photometry of the 



intensity of illumination, we find in percentages the following 

 deviations : — 







English 

 red. 



Chromate 

 of potash. 



Zinc 

 green. 



1 



In the mean. 





Carbonate 



of 

 magnesia. 



Plaster 



of 

 Paris. 



in p.c. 



5'3 p.c. 



5'8 p.c. 



61 p.c. 



101 p.c. 



4-6 p.c. 



6' in „ 



5-7 



4-6 



5-0 



140 



4:9 



1 G" in „ 



2-2 



2:5 



,7 



2-5 



1-8 



Considering that the probable error in the investigations 

 is 2 per cent, we see from these tables that the law of the cos i is 

 not valid. The values of the angles of incidence for constant e 

 are in the mean smaller than Lambert's, a result which 

 Seeliger and Wiener have also found. The law of the cos e is, 

 on the other hand, perfectly correct, whilst Lambert's law in 

 totality naturally shows great deviations caused by the cos i. 

 Hence a law of illumination can never be symmetrical in 

 reference to i and e. 



With respect to the results of other observers it seems 

 somewhat surprising to find the general validity of the law 

 of the cos €. But examining the results more closely one 

 finds that the observations come the nearer to the law of the 

 cos 6 the more matt the plates were or, what is the same thing, 

 the less normal reflexion they showed. Well, it seems to me 

 at the first start hardly reasonable to expect that plates which 

 reflect the light normally can follow a law which does not 

 take this into account at all ; it is clear that the plates must 

 be really matt if one wishes to deduce a theory from obser- 

 vations of them. This makes it plain why I took so much 

 trouble to obtain matt plates from the powders, and the 

 results show that by this method of pressure one is able to 

 get them in the required condition. It is worthy of regard 

 that it has not been possible yet to prove Lambert's law 

 theoretically as Zollner and Seeliger have shown. But that 

 is in my view no reason why it should not exist. As shown 

 in the beginning, the very sensitive objective observations in 

 the analogous case of radiant heat likewise confirm the law 

 of the cos e, whilst the law of the cos i is also found to be 

 not correct. 



Hence von Lommel's law is not valid. There are a number 



