Thus the nearest point to the land at which a sample was taken was 4 kilom., where the sea had a depth of 75 metres, and hence the chance of any disturbance in the normal bacterial contents of this seawater from accidental contamination from the shore is very slight; and even if any such source of error should arise, it would at most be only likely to take place during or after violent storms blowing from the coast.

Examination of sea-mud revealed the presence of very large quantities of bacteria. It was found, however, that the numbers steadily diminished up to a certain point with the greater depth at which the samples were abstracted, this being especially noticeable in the immediate vicinity of the coast. After 250 m. and up to 1,100 m. no further important reduction was observed.

Number of Bacteria found at various depths in 1 c.c. of sea-mud and 1 c.c. of sea-water respectively (Russell)

Russell states that these higher figures must not necessarily be regarded as indicating a larger number of bacteria as being present at the greater depth, but are to be attributed rather to slight local variations, as the samples were collected over as wide an area as possible.

As regards the varieties of organisms present, it was ascertained that more than half appeared only to belong to this mud, and were not discoverable in the sea-water itself; three individual microbes in particular were especially characteristic of this mud, as much as 35 per cent. of the total number of colonies found on

gelatine-plates consisting of these three forms (see pp. 454-456).

1

Russell has more recently extended his investigations to an examination of the sea-water and mud in the vicinity of Wood's Holl, Massachusetts. The number of microbes present in these more northern and cooler waters was markedly less at this point than in the Mediterranean. The slime from Buzzard's Bay yielded an average of 10,000 to 30,000 bacteria in 1 c.c., which Russell says represent but a small fraction. of those present in the Mediterranean mud at equal depths. Here again two species were found to be specially prevalent in the water, together with two or three other forms occasionally met with. The mud also contained these two prevailing water forms, but another form, an indigenous slime bacillus, occurred in such large numbers as to make up from thirty to fifty per cent. of the whole quantity present.

Samples of mud were also obtained, about 100 miles from the shore at the depth of 100 fathoms, on the edge of the great continental platform skirted by the Gulf Stream. These samples are the farthest from land that have ever been bacteriologically examined, and bacteria were present in large numbers; moreover, the two prevailing species present were identical with those obtained near the shore at Wood's Holl.

Russell mentions that the Cladothrix intricata (see p. 517) was only rarely met with, whereas in the Mediterranean mud it was frequently found.

On comparing these results with those already referred to for fresh-water lakes, it will be seen that the distribution of bacteria in the two cases is substantially similar, both the lake and the ocean at a

1 Botanical Gazette, vol. xvii. p. 312. 1892.

distance from land being characterised by poverty in bacterial life.

Having now obtained some idea of the bacterial contents of various waters, we must direct our attention to a consideration of some of the numerous methods which may be adopted for removing them.

117

CHAPTER V

THE PURIFICATION OF WATER FOR DRINKING PURPOSES

Sand Filtration.-It is obvious that, with the information which the bacteriological investigation of water has furnished us, the subject of water-purification must be approached from an entirely novel point of view; that, whereas formerly the chemical standard was the only one which could be appealed to as a guarantee of the suitability or not of a water for domestic supply, we have now a far more delicate test as to the efficiency of the purification processes employed in the biological examination to which a water can be submitted. Thus, perhaps, a concrete example will most clearly illustrate how the purification of water must now be regarded. Supposing that a water, derived from a source which is altogether unimpeachable as regards contamination with animal matters, is yet so highly impregnated with vegetable constituents as to be unpalatable, the question will arise how this water may be treated so as to free it from this blemish and render it suitable for drinking purposes. In a case of this kind it is obvious that chemical purification will be of paramount importance, whilst the removal of organic life from the water will be of less pressing consequence. On the other hand, if water which is known to have received sewage matters (and the entire exclusion of such from supplies drawn from rivers is practically impossible) is to be supplied for dietetic use, and if this water, as is so often

the case, is not objectionable on account of the absolute quantity of organic matter, as revealed by chemical analysis, which it contains, but only because of the suspicious origin of a part of this organic matter, then it is evident that in the purification of such water the point to be taken primarily into consideration is how the organic life it contains can be reduced to a minimum.

In estimating the value of such processes of purification, it has hitherto been customary to assume that those processes which effect the greatest chemical improvement in water may also safely be considered to be biologically the most excellent; and, conversely, that those processes which effect little or no reduction in the proportion of organic impurity are not calculated to be of any service in removing organized matters.

The following chemical analyses of river-water before and after sand-filtration will sufficiently explain how, for example, the process of sand-filtration found comparatively little favour as long as the chemical analysis was the principal basis on which a judgment as to the hygienic value of water-filtration could be formed.

Chemical Analysis of Water of River Ouse before and after
Sand-filtration (Percy Frankland)

Results of Analysis expressed in parts per 100,000