Ward . — Thames Bacteria , III. 
243 
the type of plate-colony is the same, but the zoogloeas are 
more condensed, and the colour more golden or chrome- 
yellow than No. 23. 
The chief differences in stab-cultures were the root-hairs in 
early stages, and the deep chrome-yellow yolk-like veil ; the 
streak-cultures also liquefied more rapidly. 
The agar-cultures were also much brighter yellow in hue, 
but the growth on potato was feebler than the type. 
In broth, the brighter colour and absence of veil dis- 
tinguish it. 
It seems likely, in view of all we know, that the more 
condensed growth on the plates and the deeper colour go 
hand in hand. If this is so, there may be no sufficient reason 
for removing No. 17 from the Proteus type. I have already 
shown that No. 17 and Zimmermann’s B.ochraceus are closely 
allied, and they must go together. 
That these forms are closely allied to Zimmermann’s 
B. radiatus — see Nos. 23 and 7 above — will scarcely be 
denied ; and this being so, the position of all these forms in 
the Proteus type seems justified. 
B. arborescens (Frankl.). Figs. 54-62. 
This form is by no means rare in the Thames, and was 
described by Frankland in 1889 *, but it is evidently 
merely a form-variety of the group we are considering, as the 
following study of it shows. 
It exhibits very neatly the connexion between the liquefying 
and non-liquefying conditions, and I have no doubt it can be 
modified much more than I have succeeded in modifying it. 
There can be little hesitation in connecting it with Zimmer- 
mann’s B. radiatus and B. ochraceus on the one hand, and 
with the Proteus type and allies on the other. 
Habitat. Fairly common in the Thames, especially in the winter. 
Morph. Non-motile rods, about 5 or 6 /x long by o-5~o-6 [i broad, 
often in pairs, or associated in irregular chains. No traces of spore- 
formation observed. 
Zeitschr. f. Hygiene, B. VI, p. 379. 
