DISTRIBUTION OF B. COLI COMMUNE 
1 5 
similarly examined. The plates of the analyses of January 17th were full of 
colonies, contrasting with those made from the Vyrnwy waters, which were, in 
many cases completely sterile. On this occasion the B. coli, when present, was found 
in greater numbers, about 10 per c.c. in some cases, though definite enumerations 
were not made. On the other occasions the numbers of B. coli when found were 
as low as was the case in the Vyrnwy waters. Fifty-three samples were analyzed ; 
B. coli was found to be absent in thirty-eight cases, and present in eleven cases ; 
in four cases its presence was doubtful (see Table Va). Here, as was the case 
with the rivulets of the Vyrnwy watershed, many of the streams shown to contain 
B. coli have long been suspected of pollution, and in the other cases there is distinct 
evidence of the possibility. The water of the Rivington Reservoir was also ex- 
amined in quantity, about i\ litres being used for analysis. Water from the 
surface of the reservoir showed the B. coli to be absent in 119 c.c, while water 
from the depth contained the bacillus to the extent of 27 organisms per litre 
(approx.). These analyses were made one and three days respectively after the 
samples had been taken. 
The streams and drains flowing into the River Severn near Shrewsbury were 
also analyzed (see Table Vb). Seventeen analyses were made, and in twelve cases 
B. coli was absent in 1 c.c; in the others it was present to a greater or less degree. 
Number 11, which contains B. coli commune to the extent of 19 per c.c, enters the 
River just below Wroxeter, and, from the map, it seems probable that it would be 
contaminated with domestic sewage. Number 15 is a stream which receives the 
sewage from Atcham Union Workhouse. There are considerable difficulties in the 
way of finding out exactly the previous history of such streams and drains as these ; 
probably all those containing B. coli have received some definite pollution. The fact 
that the majority of the land drains are free from B. coli is, I believe, contrary to a 
statement that is often made, that this bacillus is to be found in quantity in all 
cultivated lands, and hence in their drainage. Of course, such water has been 
subjected to a very fair filtration before it enters the drains. 
(g ) Sand of Filters used for filtering Drinking Water, — The water of the Vyrnwy 
Lake is filtered through sand beds before being used for drinking purposes. It 
contains the Bacterium coli commune, though to a very small extent. All attempts 
to detect the presence of the bacillus in the filtered water have failed, even when large 
quantities have been used for analysis (see below). Hence it seemed probable that 
the sand beds used for filtering might retain the bacillus, and this has proved to be 
the case, confirming other experiments on the efficiency of sand as a medium for 
filtration. The sand of the filter beds was analyzed in the same way as the soils 
previously described ; in these cases a dilution of 1 gram in 10 c.c. of sterile water was 
invariably used. The details are given in Table VI. Out of 15 samples analyzed, 
B. coli was absent in .1 gram in 8 cases and present in 7. In one of these cases 
