298 BACTEHIOLOG Y 



seen, 1 when examined in water by the micro- spectral objec- 

 tive, to make their way to the part furthest from this end. 

 Janowski exposed cultures under screens of various chemical 

 solutions and aniline dyes, and found that no action took 

 place under brown or yellow, whereas solutions of fuchsine 

 (which transmits violet rays), gentian violet, and methyl 

 blue had little more effect than colourless fluids. Confirm- 

 atory results have also been obtained quite recently by 

 Marshall Ward, 2 who repeated his experiments with the 

 anthrax plates, covering the stencil letter with pieces of 

 glass of various colours, as well as with screens of saturated 

 solutions of potassium bichromate and of ammoniated 

 cupric oxide (which cut off respectively the violet and red of 

 the spectrum as far as the line 6), and found that no effect 

 was produced on the spores by exposure to sunlight behind 

 ruby, olive, orange, or green glass, nor behind the bi- 

 chromate solution (when sufficiently concentrated), while 

 4 hours' exposure behind violet and blue glass, and a much 

 thicker cupric oxide screen, was usually sufficient to kill 

 them. Less pronounced effects were produced by diffuse 

 daylight. The bactericidal action of light, therefore, seems 

 to depend on the more refrangible rays in the violet half of 

 the spectrum, which produce their effects whether the red- 

 yellow rays are transmitted or not. 3 



The mode of action of light. That the effects produced 

 are really due to the rays of light and not to heat is shown 

 by the fact that they also take place under water (Buchner) 

 and behind a screen of alum solution (Janowski), that a ther- 

 mometer at the surface of the glass never registered above 



1 Engelmann, quoted without reference by Woodhead (Bacteria and their 

 Products, 1892, p. 201). 



- Boy Soc. Proc., vol. liii. No. 321, p. 23. 



3 Prof. Ward states (May 25) that he has made great advances in his 

 experiments with chemical screens and with the spectrum, but has elicited 

 no contradictions to his earlier results, as already published. 



