Report on the Bacteriology of Water. 269 



It hardly needs pointing out that these are merely so many stages 

 in the progressive segmentation into bacilli, each bacillus being 

 therefore a cylindrical rodlet, from 3 to 5 or 6 M long by about 1'75 /t 

 broad, and with rounded ends (see figs. 9 13). 



Before this complete segmentation into bacilli, however, which 

 only occurs, as a rule, very late in the life-history, the septate fila- 

 ments often break across here and there at a septum, and so we have 

 shorter filaments, each with twenty, fifty, or a hundred or more septa. 



As I shall have to recur to these matters when tracing the develop- 

 ment, however, no more need be said here further than that fila- 

 ments and rodlets alike are usually quiescent, even in the liquefied 

 gelatine,* and no cilia can be detected at any time; for the very 

 slow growth-movements to be described later would hardly come 

 under the head of motility in the sense ordinarily used in bac- 

 teriology. 



Streak-cultures on gelatine at 15 C. result in very pretty and 

 characteristic growths. A rapid extension of the tresses and fila- 

 ments above described is effected from the streak all over the surface 

 of the gelatine, so that in three or four days the white mycelium- 

 like membrane covers nearly the whole area (fig. 4), and presents 

 some resemblance to the diagram of the digestive and blood systems 

 of certain worms. Meanwhile the gelatine begins to soften and 

 liquefy, and on the fifth day the film, more or less broken up, is 

 floating on the liquid. In ten days nearly all the gelatine is liquefied, 

 and flocks of broken film float on and in it. The liquid does not 

 show any noticeable turbidity at any time. 



Stab-cultures in 10 per cent, gelatine at 15 C. are even more 

 beautiful. From the white axis formed along the line of puncture, 

 radiating silky-looking filaments branch out horizontally towards 

 the walls of the tubes, and by the second or third day the culture 

 looks so like a root, with its silky root-hairs radiating around, that 

 this stage would seem best to justify the German name Wurzel- 

 bacillus : it resembles strongly some cultures of Bacillus anthracis. 

 On the fourth day liquefaction of the gelatine has definitely begun 

 at the top, and the " root-hair system " now looks more like an 

 inverted fir-tree (to use Fraenkel's simile), and this resemblance 

 becomes more and more perfect as the culture gets older (see figs. 5, 

 6, and 7). 



The liquefaction of the gelatine progresses slowly from above 

 downwards, and the felted yellowish-white membrane floating on 

 the top gradually breaks into flocculent patches. Even after three 

 weeks the liquefaction has only extended about one-eighth of the 

 distance from the top of the gelatine, but the fir-tree resemblance is 



* A slow movement of isolated segments breaking off at high temperatures (about 

 35 C.) must be regarded as abnormal, as will be seen later on. 



