DISTRIBUTION OF B. COLI COMMUNE 123 
Table III). The dampness may have been caused by artificial watering of the street 
in question, or by the constant trickling of water from a fish shop serving to keep a 
gutter moist, as in No. 14. 
Whether a moist substratum would serve to resuscitate B. coli, which had 
apparently died out, or whether in such dry samples the bacillus was quite dead, was 
answered in the following manner. The dilutions of May 31 (see Nos. 26-31, 
Table III) were kept for three days and then analysed again. The results showed 
that in the cases in which B. coli was absent on May 31, it was also absent after 
the sample had been thus kept wet and favourable for the growth of the bacillus ; 
in those cases in which it was present on May 31, its numbers were much reduced 
on June 2. One must conclude that B. coli, constantly being thrown upon the 
roads, is able to survive in dry weather for a short time only (No. 33) ; if, however, 
the weather is wet or the road kept damp artificially, it can live much longer. 
These facts suggested the enquiry whether sunshine or drying was the 
important factor in the short life of B. coli in nature under these conditions. 
The effect of drying has been studied by Walliezek (Centralbl. f. Bakt. X_F, S. 949), 
in whose experiments 17 hours' exposure to drying over sulphuric acid killed 
almost all the bacteria on a piece of filter paper that had been dipped in a 
culture of B. coli and then dried ; a similar effect was obtained after 45 minutes' 
drying in a vacuum. Billings and Peckham (Science, Feb., 1895, Abstr. Centralbl. 
f. Bakt. XIX, S. 244), found that threads moistened with a broth culture of B. coli 
and then dried, showed living bacteria for a period of 152-229 days, while sunlight 
alone proved fatal in 3-6 hours ; they concluded that death by sunlight is on no 
account to be attributed to drying. Buchner (Centralbl. f. Bakt. XI, S. 781), also 
showed the importance of sunlight in killing B. coli ; in one experiment 100,000 
B. coli suspended in water and exposed to sunlight were all found dead after one 
hour. 
The experiments of Wittlin (Wiener Klin. IVocbenschr., 1896, No. 52, 
S. 1,229), however, indicate that drying must be a factor in the death of B. coli in 
nature, as important as sunshine, if not more so ; and the experiments I am about 
to describe confirm his observations. 
An emulsion was made with a twenty-four hours' culture of B. coli on agar and 
a little water, and this was used to moisten some sterilised earth, which was then dis- 
tributed into six Petri dishes. The soil in all of these was analysed for B. coli in the 
way already described ; in the subsequent analyses, where the earth was really wet, it 
was dried before weighing on sterile filter paper. This method, I think, makes the 
numbers of the wet samples rather lower than they should be, and renders the 
differences a little less striking than they really are. After analysis, water was added 
to three of the six samples, and these were kept thoroughly wet throughout the 
course of the experiment, the other three were allowed to dry. One of each was 
