1887.] Mr A. W. Hare on the Purity of Water . 
307 
importance to inquire if their relation to one another can he more 
closely analysed, and whether specific features of the one factor 
are necessarily related to a certain constitution of the other. A 
well-established case of such a relationship may be here cited in 
illustration, and as indicating the direction in which the present 
inquiry is to proceed. The Beggiatoa alba , a well-known fila- 
mentous organism, is an inhabitant of warm sulphur springs, in 
which it carries on a definite analytical function, decomposing 
sulphur compounds in solution in the water, and giving off sul- 
phuretted hydrogen. That such facts are of frequent occurrence 
is shown by the work of Pasteur in the case of the Torula group ; 
while the elaborate researches of Duclaux exhibit functional 
specialism in this respect carried to an astonishing degree of com- 
plexity. It is not then beside the mark to press the analysis of 
aquatic microbes one stage further, and to inquire whether, in 
addition to the fact of organic contamination predicable from their 
presence, there is not a possibility of recognising definite species, 
associated with special forms of organic material, particularly such 
as are derived from sewage matter. In making this attempt an 
initial difficulty is met with in the fact that a complete classification 
of Bacteria according to their function is not yet made; their 
provisional division into zymogenic, pathogenic, and chromogenic 
forms, though of great dialectic convenience, is of no value in practical 
questions, for each division is in part overlapped by portions of the 
other two. A new classification must therefore be attempted to suit 
the conditions of this inquiry ; and that which specially commends 
itself as at once practical and well adapted to our purpose is based on 
observing the nature of the pabulum on which various species subsist; 
thus dividing them into groups according to the complexity of their 
assimilative processes, as higher groups of organisms may be classi- 
fied according to their carnivorous or herbivorous habits of life. In 
the case of the Bacteria this classification can rest on no such 
obvious diversity of function, for they are distinctly omnivorous ; 
but there are important differences in the composition of the organic 
matter in water, corresponding to the different sources from which 
it is derived, and it is not unreasonable to expect that analogous 
differences will be found to obtain amongst the species of microbes 
associated with it. Two chief sources of organic material in rivers 
