Report on the Bacteriology of Water. 



325 



two and a half hours, and distinctly more so in both cases than where 

 glass was employed in addition. 



IV. 



The following series of experiments (Table <D) were made in con- 

 tinuation of the foregoing, and the description of glass screen em- 

 ployed (3rd column) refers to the table on pp. 328 329. 



The chief feature of novelty is that I here used the same plate 

 for different screens, as follows. A thick, opaque cardboard screen 

 was prepared as large as the plates, and this screen perforated with 

 four or five circular or square windows. Over each hole a piece of 

 the coloured or other glass to be employed was then cemented, and the 

 whole held in contact with the plate to be exposed, by elastic bands. 



The advantage of this proceeding was that I could check the pre- 

 ceding results, to see if any erroneous conclusion had resulted from 

 my using different plates. 



V. 



In order to investigate more in detail the action of the decomposed 

 sunlight on the spores, I made a series of glass screens of the nature 

 of water cells, or reservoirs, as follows. A number of circular, flat, 

 indiarubber packing rings, about a quarter of an inch thick and three 

 inches internal diameter were obtained, and a small piece cut out of 

 each ; then a thin, plain piece of glass was cemented to each side of 

 the now incomplete ring, thus forming a reservoir with two large 

 parallel glass ends, the sides being formed of indiarubber. I found 

 that by carefully cementing the glass with gold size, it held very 

 well, at least for two or three experiments, and could easily be re- 

 cemented if necessary. 



FIG. 1. 



These glass cells were filled with the coloured transparent solu- 

 tions to be referred to (see Table E), and then placed over the 

 exposed letter on the prepared plate as described in my previous 

 paper (' Proc. Roy. Soc.,' vol. 52, p. 393), being held in position by clips 



