Notes. 
345 
In order to discount as far as possible variations in the salt-content of the 
tap-water due to dilution by rain, &c., sets of experiments were made over a period of 
four weeks. There appeared to be no substantial variation.from this cause. 
Titration curve of boiled tap-water with carbonic acid of Ph 5 ± 0*05. The curve repre¬ 
sents the mean of nine experiments. Probable error of the mean, less than 2 per cent, of the 
mean, except for points corresponding to Ph 8*69 and 8.08, where it is 5 per cent, of the mean. 
Indicators : thymol blue (r in 1,000) and phenol red (1 in 10,000), three drops of each to 10 c.c. 
water. Temperature i6°C. + 1. 
The results may be expressed by the above curve, which is constructed 
from the mean of nine experiments. It will be seen that the buffer action is least 
