1903.] Bactericidal Action of some Ultra-violet Radiations. 127 



cadmium. Visually the lines in the ultra-violet region can be observed 

 by projecting the spectrum on to a fluorescent screen of sulphate of 

 quinine. We, however, obtained photographs of the cadmium spectrum 

 under precisely the same conditions as those under which the bacterial 

 plates were obtained, and were thus able to identify the lines occurring 

 in the photograph, with those on the bacterial plates. As these lines 

 are known, their wave-length is at once obtained by comparing the 

 photographs with a chart on which the wave-lengths are indicated. 

 The photographic method, therefore, reduces the chance of mistakes 

 which might arise from personal error in observation. It, therefore, 

 appears that relatively the action of other portions of the spectrum is 

 negligible compared with the activity of this portion, although it is 

 probable that when using white light there is a slight action which 

 extends over the whole spectrum. Neither the extreme ultra-violet 

 rays nor those nearest to the visible violet appear to be active. The 

 affected portion of the bacterial plates corresponded with a photograph 

 taken of that portion of the spectrum, and it was possible to identify 

 the nearly sterile lines on the plates with those known to exist in the 

 ultra-violet spectrum of carbon. 



In view of this result we were led to experiment with the arc spectra 

 of various metals, such as iron, cadmium, silver and aluminium. The 

 results with these entirely agreed with those obtained with carbon, 

 except that the action is greater in proportion to the number and 

 intensity of the lines occuring in what we may call the bactericidal 

 region. 



It, therefore, appeared that an electrode composed entirely or partly 

 of iron, should be more actively bactericidal than a carbon one, and 

 this we found to be the case. A convenient form we found to be one 

 in which in the case of the positive electrode, we removed the soft 

 carbon core and substituted for it a mixture consisting of the particular 

 metal desired and sufficient carbon, in the form of sugar, to prevent 

 the core from dropping out. Using a positive electrode so prepared 

 and an ordinary carbon negative electrode, the effect on Bacillus coli 

 communis contained in a hanging drop was as follows : 



Distance of hanging drop from arc, 10 cm. Current used 11 amperes. 



Ordinary carbons Organisms killed in 30 mins. 



Iron 15 



Cadmium 15 



Aluminium ,, 25 ,, 



The organisms were proved to be killed by failure to obtain sul>- 

 cultures, but the motility was arrested long before the death of the 

 organism. 



The light was allowed to pass through a metal cylinder, closed at 



