Report to the Water Research Committee. 



417 



density, &c, of which is changing all the time, and since every 

 variation in temperature, light, air-supply, &c, affects the matter 

 also, I find abundant causes for the variations which otherwise 

 appear so astonishing. It results, therefore, that the attempt te» 

 determine species of bacteria by ordinary macroscopic methods leads- 

 to difficulties of the same kind as would be met if we tried tc» 

 differentiate species from the marks presented by masses of trees ia 

 forests from a distance — say in a balloon. A forest of a given 

 species of tree would appear very different at different seasons, and 

 according to its age, the kind of soil, climate, and so on, and the 

 treatment it had received previous to planting. 



In order to emphasise my own impressions of the specific value of 

 the forms isolated, and at the same time leave it clear what are the 

 differences between the various forms or varieties themselves, I have 

 grouped them into types. Each group contains a type which I 

 regard as probably a species, of which the other forms included are 

 varieties. These varieties are fully described and figured, and would 

 certainly be regarded as species in most books consulted. The 

 reasons for grouping them as I have done are fully discussed with 

 the literature under each. Here, of course, 1 can only give a very- 

 brief summary of the groups as follows : — 



Group I is made to contain a form which appears to be identical 

 with the Bacterium ureoe. of Jaksch, and forming curiously character- 

 istic dull stearine-like plate colonies. It was not uncommon in the 

 river, but the want of salient features rendered it comparatively 

 uninteresting, and no prolonged study of other varieties of it was- 

 undertaken . 



Group II was constituted for a type which in certain cultures- 

 developes an intense violet colour exactly like that of a strong 

 solution of gentian violet. It was not common — at any rate in the 

 typical form — but it was easily cultivated, and two varieties were 

 isolated and kept under observation for many months. 



The type accords best with Jolles and Eisenberg's B. membrana- 

 ceus amethyst inns, and forms the deep blue violet zoogloea-membrane 

 described for that species, but three years' study of the variations in 

 culture convinces me that several violet bacilli described in the 

 literature are identical with this, or at least cannot be separated by 

 the characters given. 



A marked tendency to lose its power of forming pigment, and to. 

 grow as a pure white form, characterises this bacillus. It is the 

 species with which I obtained the bacterial photographs exhibited in 

 1894, and is very sensitive to light. The spectroscopic characters of 

 the pigment were determined, and the bacillus itself cultivated 

 under high powers. 



Group III comprises the common B. fluorescein liquefaciens, very 



