390 



Profs. Percy Frankland and Marshall Ward. 



Before removing to the incubator, at 24 C., I examined both care- 

 fully. No trace of a rodlet could be found in the water culture, but 

 30 or 40 filaments from 200 to 300 and even 450 p, were abounding 

 in that over bichromate. 



At 7 P.M., re-examination showed immeasurably long filaments in 

 the bichromate culture, whereas only three filaments 58, 96, and 81 /* 

 long respectively could be found in the water culture. 



I think this must be regarded as proof that light-action occurs 

 under the above conditions of exposure ; and there can be little doubt 

 that with a more powerful lamp perhaps aided by quartz con- 

 densers, &c. effects more nearly approaching those with sunlight 

 could be obtained. 



Antagonistic Action of a Swarming Bacillus. 



The following experiment illustrates very clearly the effect of in- 

 troducing a rapidly growing, strongly aerobic bacillus to compete for 

 oxygen, &c., with the schizomycete in question. 



I selected a well grown culture of B. ramosus, and introduced into 

 the broth-drop a few rodlets of perhaps the most rapid growing 

 liquefying bacillus yet isolated from the Thames. 



This form belongs to the group commonly known as B. termo, and 

 its peculiarities will be described in due course. I call it, for the 

 present, form /3. It is a short, stout rodlet, about 1*5 to 2ytt x lyu, 

 actively motile, a very rapid grower, and exceedingly greedy of 

 oxygen. 



At 8.50 A.M., a portion of the B. ramosus, around which the bacillus ft 

 was rapidly swarming in considerable numbers, was put into measure- 

 ments, with the following suggestive results : 



Curve 77A. 

 It doubled its length in 115 minutes at (air-temp.) 18 22 C. 



