16 THOMPSON YATES LABORATORIES REPORT 
the approximate number of B. coli per gram of sand was as high as 2,055, while in 
another place on the same bed it was 20. 
Similar analyses were made of the sand of the filters at Rivington, which filter 
the water of the Rivington Reservoir. The sand of three of the beds was examined ; 
in all 14 analyses were made, and the B. coli was found to be absent in each case 
(see Table Via.) 
(h) Drinking Water. — Eighteen samples of filtered water from the filters a' 
Rivington were examined, and in every case B. coli was found to be absent in 1 c.c. 
(see Table VII). The tap water of the laboratory was also found to contain no B. 
coli in 1 c.c. This water was also analyzed in quantity on two occasions (March 9th 
and March 13th, 1900), about 2^ litres being concentrated by means of a Chamber- 
land and Berkefeld Filter respectively, and a portion of the concentrated water used 
for analysis. In the one case B. coli was found to be absent in 900 c.c, and in the 
other in 823 c.c. Of course, the Liverpool water supply is unusually excellent. 
On February 26th the dead ends of the pipes supplying certain streets in 
Liverpool were cleaned out by flushing with water, and samples were taken at certain 
intervals of the dirty-looking water which came out (see Table VII<?). Sixteen samples 
were taken and analyzed, and B. coli was found to be absent in every case. 
The drinking water of Shrewsbury was analyzed (November 18th, 1899) and 
showed the presence of B. coli. The water supplied to the town by conduits from a 
distant source contained 2 B. coli per c.c. The river water, filtered, is also largely 
used in the town, and contained 17 B. coli per c.c; this water is of an undeniably 
polluted origin. The water of the river, at the point where the town supply is drawn 
off, has on several occasions been analyzed and has always contained the B. coli. 
(/) Grain and other Foodstuffs.- — These analyses were made in order to institute 
a comparison between the respective distribution of B. coli commune and B. enteritidis 
sporogenes. The duplicate analyses were made by Dr. Balfour Stewart, who also 
investigated the significance and comparative frequency of the latter organism (see 
this volume, pp. 31 and 34). In all 19 samples of grain were examined, including 
English and foreign wheat, both before and after malting, rice, oats, &c The 
analyses were made thus : — Different quantities of the grain were placed in a sterile 
flask, and well washed with 10 c.c. sterile water, 1 c.c of this being used for plating. 
B. coli commune was found to be absent in every case (see Table IX). Some samples 
of crushed grain were also examined ; no particular quantities were taken, a little was 
well shaken up with 10 c.c. sterile water, and 1 c.c. plated as before. B. coli was 
absent in crushed malt, oat husks, bran, and also in a sample of flour, which was the 
coarsest quality sold. 
The comparative infrequency of the occurrence of B. coli commune in foods 
may also be seen from Table IX, which contains a summary of food analyses made in 
the Thompson Yates Laboratories for the Liverpool Corporation, extending over 
