62 



Messrs. H. Hartridge and A. V. Hill. 



Preliminary observations with lens preparations made in this way showed 

 us that the absorption bands in the infra-red corresponded very closely in 

 position to those of water. On closer comparison, however, we found that 

 there was in addition in the lens preparation what appeared to be a more 

 general absorption which gradually increased in amount as one passed toward 

 the visible spectrum. Fig. 2 is typical of the results we obtained with 

 several different lens preparations. Several explanations occurred to us to 



1 1 1 



7500 A.U. 10,000 12,500 



Fig. 2. — Comparison of Amount of Infra-red Energy of Different Wave-length 

 transmitted by Lens and by an Equal Thickness of Water. 



Swing of galvanometer vertical. Wave-length horizontal. 



account for these results. We first supposed that there was some substance 

 present in the lens with a very diffuse absorption band. This, however, would 

 not fit in with the fact that the absorption band, extending as it did to 7500, 

 would be seen in the visible spectrum and would therefore cause the lens to 

 appear greenish in colour. Our second theory was that in spite of the special 

 condenser system, described above, we were getting the effects of the difference 

 in dispersion between the visible and infra-red rays, so that, while the infra- 

 red rays in the case of the figures mentioned above were properly focussing 

 on to the slit, some of the visible rays crossed too early and were lost. This 

 explanation too had to be abandoned for several reasons. In no case were we 

 able by changing the focus to get the reverse effect, i.e., the visible rays giving 

 full values and corresponding to water, and the infra-red rays giving values 

 which fell more and more away from the water curve. A considerable change 



