246 Ward . — Thames Bacteria , III. 
that of No. 23, i. e. slowly growing circular colonies with an 
ochre-yellow or leather-yellow central spot, and sometimes 
a zone of the same colour, and with very evident radial 
structure. On magnifying this, one finds densely branching 
and interwoven filaments radiating and crossing one another 
from a central zoogloea of the arborescent or serpentine type. 
The extreme tips of the irregularly radiating strands are very 
thin, and give the fimbriated appearance shown in Fig. 57 
(cf. Fig. 37). These colonies slowly liquefy the gelatine, and 
each lies for some time in its own lenticular circle of lique- 
faction. 
When, on the other hand, the colonies are developed from 
Bacilli in full vigour, e. g. after passing through broth and not 
exposed to light, they appear in forty-eight hours or so as 
serpentine zoogloeas with thin fimbriated branches from the 
ends, and at once liquefy the gelatine throughout, in the 
manner typical for a Proteus , without developing into formed 
visible colonies. This is because the Bacilli separate off and 
wander all over, and liquefaction is so vigorous that no con- 
densed growths are formed after the first (Fig. 56, and cf. 
Fig. 44). 
There is no doubt that in these experiments the form of 
colony is dependent on the conditions to which the Bacilli 
have been exposed, because the latter have their powers of 
liquefaction, and therefore of nutrition and growth, affected. 
Unfortunately it is impossible to so arrange matters that only 
one factor at a time is varied in the environment, and the 
experimental proof of the action of each individual factor is 
therefore wanting. We have here, however, ample proof that 
it is important to keep the conditions as constant as possible 
in plate-cultures. 
Sanfelice 1 had already suggested that B. a 7 'borescens is 
a variety of Hauser's Proteus , as is also B. aquatilis of the 
same authors. My studies of the yellow liquefying forms 
which I have classed as Group IX in my fifth report to the 
Royal Society 2 , suggest that Zimmermann’s B. fulvus , 
1 Lustig, 1 . c., p. 85. 2 Roy. Soc. Proceedings, Vol. lxi, p. 419. 
