420 



Prof. H. Marshall Ward. 



nearest to Adametz's B. aureus and P. Frankland's B. aurantiacus. 

 They grow very slowly, and seem to be feeble forms. From their 

 behaviour during culture and comparison with the last group, I am 

 compelled to regard these non-liquefying yellow forms as merely 

 weakened varieties of the last group ; this is borne out by the 

 tendency to liquefaction shown by some of them after being culti- 

 vated for some time, as well as by the enfeeblement of liquefying 

 power shown by some of the older cultures of the yellow Proteus 

 group. It is probable that they are simply species of the Proteus 

 group, so weakened by their prolonged sojourn in the river that they 

 can only be cultivated with difficulty in the new environment of the 

 gelatine media, &c. This being so, the whole of the series included 

 in Groups VI to X inclusive may be varieties of one or two species, 

 since I find that, great as the differences seem between the extreme 

 forms, the varieties studied present characters passing imperceptibly 

 into one another. This result is also in accordance with the fact 

 that so many yellow varieties, impossible to separate satisfactorily, 

 have been isolated from water and described by other observers. 



Group XI is made to include a colourless capsuled form, not very 

 common in the river, and reminding one of Cohn's Ascococcus. Two 

 sets of cultures of it were carried out in detail, and a very character- 

 istic feature was discovered in the plate-colonies, viz., the presence 

 of curious vermiform zooglcea masses imbedded in the colonies ; 

 these are not intruded forms, but merely growth-phases of the same 

 species. So far as can be judged, the " species " seems new. 



Group XII comprises a series of yellow capsuled bacilli, not 

 uncommon in the river, and morphologically conforming to the type 

 of Friedlander's bacillus. I cultivated five varieties of it. The most 

 characteristic feature is the formation of extremely tough zooglcea- 

 colonies, often looking like the roe of a fish. There is some variation 

 in the degree of liquefying power, and much difference in consistency 

 and other characters, but the sum of all the diagnoses leaves no 

 doubt that all the forms belong to one species. 



Group XIII comprises a series of forms which produce a crimson 

 pigment. The type is that of B. prodigiosus, and five varieties were 

 isolated and very closely investigated. Colourless varieties were 

 developed in cultivation, and the pigment again restored by suitable 

 treatment. Weakened forms, and considerable differences in liquefy- 

 ing power were also found. Koch's B. indicus, Breunig's Kieler 

 bacillus, Frank's B. ruber, DowdesweiPs B. rosaceum metalloides, and 

 Zimmermann's B. miniaceus are all identical with this form, or, at 

 least, can be matched by the varieties arising in culture. 



Group XIV comprises a series of rapidly liquefying, colourless 

 foims, very common in the river at all times, and conforming to the 

 type of B. termo as amended by Mace. I made a prolonged study of 



