Report on the Bacteriology of Water. 335 



lying spores germinate more slowly than those further away from 

 the illuminated area the stencil letter they do at last germinate 

 out, and so the previously blurred letter becomes sharp and clear in 

 outline. 



The phenomenon very ranch resembles the development of the 

 indistinct " ghosts " of letters in cases where the exposure is too 

 short, or the light not sufficiently intense, or wanting in active rays. 

 Such faint letters gradually become obliterated as incubation pro- 

 ceeds, because the spores, still alive but only retarded in develop- 

 ment, gradually germinate out to an extent so little differing from, 

 the rest that the eye fails to detect any difference. 



The retarded development of a few colonies, at a late period of in- 

 cubation, on the hitherto clear area of the exposed letter, is due to 

 similar causes, but produced in a slightly different way. It is ex- 

 tremely difficult (probably impossible) to thoroughly distribute the 

 spores in the film so that some do not shelter others from the light ; 

 consequently, when a clump of spores exists on the exposed area 

 some of the inner spores may so far escape the bactericidal action as 

 to be able to germinate out later, and I have had many experiences of 

 these cases. In fact, the chief point about a good film i.e., one 

 which develops a sharp letter after ordinary exposure is that the 

 spores shall be neither too few nor too many, and thoroughly and 

 evenly separated and distributed ; and, lastly, that the agar or other 

 medium shall not be too thick, and thus render possible the ordering 

 of long rows of spores one behind another (i.e., in rows parallel to the 

 ray of incident light) which thus shelter one another from the light's 

 action. 



These points, and some others, come out still more clearly in the 

 next series of experiments. 



vn. 



The following series of experiments were made behind superposed 

 screens, and it must be borne in mind that the light had to traverse 

 not only a double thickness of solution, but also five thicknesses of 

 glass, before reaching the spores. 



On the whole these results may be regarded as simply confirming 

 the previous ones, but I was (and still am) considerably puzzled by 

 the behaviour of the iodine and carbon bisulphide screens. In several 

 cases the plates seem to be destroyed by the light passing through 

 this medium, and for some time I was doubtful whether there might 

 not be a cumulative effect due to the action of the infra-red rays. It 

 seemed extremely probable that these rays the " dark " heat rays 

 do help to promote the bactericidal action, and I thought perhaps 

 because they accelerate the chemical changes on which the action 

 depends. 



