334 Profs. Percy Frankland and Marshall Ward. 



If we now look at the results tabulated above, it is seen tbat tbe 

 solar action is evident, thougb feeble, through dilute fncbsin, sesculin 

 and quinine, and picric acid ; wbile no trace of action occurred 

 througb potassium chroraate, chlorophyll, eosin, and strong fuchsin. 



On the other hand, the action was sharply defined where am- 

 moniacal cupric oxide or water alone was employed, and also where 

 alum dissolved in water was used. In other words, the action is 

 most pronounced when the rays transmitted are those of the blue- 

 violet end of the spectrum, bearing out the results already obtained 

 more generally. 



During the progress of the experiments above tabulated, a number 

 of other points of interest were observed. With carbon bisulphide 

 and iodine it frequently happened that no letter was obtained on the 

 plates (Expts. J, 3, Ilia, 7, 9a), but occasionally the light action was 

 recorded by the appearance of the letter (Expts. S, Illfe, BZ). The 

 fact is, the solution did transmit a scarcely perceptible amount of 

 violet rays, and since I could not discover any definite relation 

 between the times of exposure and the results, one of two possibili- 

 ties suggested itself either differences in the thickness of tbe glass 

 of the plates, or differences in the degree of clearness of the atmo- 

 sphere may account for the discrepancies. Probably both causes 

 were effective, for, of course, they both affect these violet rays con- 

 siderably. 



Another phenomenon repeatedly noticed, both in these experiments 

 and in others, was that if the exposure to a very bright sun is con- 

 tinued too long, and especially if the plate is not very accurately at 

 right angles to the direction of the rays, the light may clear the 

 plates entirely, or nearly so. This seems to be due to the reflections 

 of the light from the glass surfaces inside the Petri's dishes ; if the 

 light is very intense, or the exposure long, these reflected rays are 

 sufficiently powerful to produce effects similar to those of the direct 

 light. 



This seems to me to explain another phenomenon very commonly 

 met with. In many cases of long exposure to clear hot sunshine, the 

 first evidence of the successful light action is not a sharp well-defined 

 letter, bat a blurred clear patch, which slowly sharpens up as incuba- 

 tion goes on. 



It is evident that in such cases the action of the light has extended 

 beyond the boundaries of the stencil letter, into parts of the film 

 really not exposed to the direct incident rays. I explain this as due 

 to the reflection of some of the rays from the glass surfaces in the 

 interior of the plate. 



These reflected rays are not sufficiently intense to complete the 

 bactericidal action, they only inhibit the organism more or less, or at 

 least leave many spores still alive ; consequently, while these out- 



