1895.] affecting Bacterial Life in Thames Water. 443 



the microbes were much more numerous than even in the previous 

 December. In March, whilst the temperature remained practically 

 the same as in February, the number of microbes per cubic centi- 

 metre fell from about 14,000 to about 3.700. Again, in April, whilst 

 the temperature was still moderate (11'6 C.), the number of 

 microbes fell still further to less than 2,000 per cubic centimetre. 

 In May, June, July, and August, the temperature was high and the 

 number of microbes uniformly small, but in September there was a 

 great reduction of temperature, but accompanied also by a consider- 

 able diminution in the number of microbes. In October there was 

 again, as in February, an enormous development of bacteria whilst 

 the temperature was only slightly lower than in the previous month, 

 when the number was remarkably small. In November again, with 

 a much lower temperature, there was a reduction in the number of 

 microbes from about 14,000 in the previous month to about 1,750 per 

 cubic centimetre. During the following months of November and 

 December, 1893, and January, February, and March; 1894, whilst 

 the temperature remained low and nearly constant, the microbes 

 were sometimes as low as 1,789, and sometimes as high as 56,630 ; 

 but, during the following months April, May, June, July, August, 

 September, and October the increase of temperature and diminution 

 of microbes, and vice versa, marched very much pari passu. In 

 November, however, with no alteration of temperature as compared 

 with the previous month, the number of microbes increased from 

 3,800 to 20,080 per cubic centimetre, whilst in December there was a 

 diminution alike in temperature and microbes. 



Thus it is evident that, although coincidences between a high 

 number of microbes and a low temperature are not wanting, some 

 other condition entirely masks the effect of temperature. 



The next table and diagram institute a comparison between the 

 number of microbes and the hours of sunshine to which the water 

 has been exposed. The diagram is constructed on the same lines as 

 No. 1. 



Reference to the table and diagram shows that, whilst during the 

 months May to November, 1892, there was an enormous variation in 

 the amount of sunshine, namely, from 23^ hours in June to none in 

 November, there was practically no corresponding variation in the 

 number of microbes. In December of that year, there was much 

 more sunshine than in the previous month, but the microbes, instead 

 of being diminished thereby, increased from less than 2,000 to over 

 11,000 per cubic centimetre ; and, although there was no sunshine on 

 the three days previous to the taking of the sample in January, 

 1893, the number of microbes actually decreased about 3,000 per 

 cubic centimetre. In the following month of February, however, 

 about 14,000 microbes per cubic centimetre coincided with but little 



