490 
Proceedings of the Royal Society 
Table I. It may be useful here to allude to a similar investigation 
published in the Arbeiten aus dem Kaiserlichen Gesimdheitsamte, 
Band I. Berlin, 1885, on the water supplied to Berlin. The 
cultivations were made daily from July 1884 to March 1885. The 
mean of the numbers of colonies raised from unfiltered river water 
(Stralauer Works) was 1435 per gramme, the maximun being in 
October 3251, the minimum in February 685. Filtration altered the 
water very considerably, the mean of the filtered water being 107 
(maximun 277, and minimun 21). 
Lake water (Tegeler Works) when unfiltered gave a mean for the 
nine months of 441, the maximum 890 falling in March, the 
minimum in August. The mean number of colonies in the filtered 
lake water was 51, and the extremes were 14 and 121. 
It is therefore clear that the filtered Berlin river water is, so far 
as bacterial life goes, purer than the filtered London river water, 
the mean number of colonies per gramme being for the West 
Middlesex Company 468, and that of the Grand Junction 
261.* 
With regard to other waters, I give in the following tables 
examples of good, indifferent, and polluted waters, from which 
I think it may be concluded that, irrespective of the $peci^ 
kind of micro-organism in a water, any number of colonies 
which exceed 1000 per gramme is indicative of considerable im- 
purity. 
Summary . — To briefly sum up the facts already detailed — 
(1) Impure liquids, such as sewage, pond and canal waters contain 
hundreds of thousands or even millions of micro-organisms per 
gramme. 
(2) In canals and rivers the number of micro-organisms at or near 
the bottom, is greater than at or near the surface. 
(3) In sea water the conditions are reversed, the number of 
organisms being greater at or near the surface, 
(4) A good water should not contain more than 1000 colonies of 
micro-organisms in every gramme, f 
* My numbers for the London waters have been hitherto in excess of the 
one or two other observers working at this subject. I always cultivate the 
water twenty-four hours after collection, during which period of rest there may 
be some considerable increase, especially in summer. 
t That is, when cultivated within twenty-four hours after collection. 
