240 
length and becomes stronger. The following experiments 
corroborate this. New Orleans cotton was treated with soda 
solution sp. gr. T250. Twenty hairs being measured gave a 
mean length of O'SoT inch, being a contraction of O’ 139 inch 
on the mean of twenty hairs measured before treatment. 
The mean strength of 10 hairs is 154* 1 grains against a 
mean strength of 1 38* 1 grains before treatment. 
The reliability of these results is discussed by the author. 
The experiments in detail show wide discrepancies, and the 
maximum and minimum of a series are often at a considerable 
distance from the mean. This is a difficulty inherent to the 
subject, and can only be overcome or lessened by multiplying 
the experiments. Ten experiments seem to give a reliable 
mean, for in making twenty or thirty, the first, second, and 
third tens give nearly the same mean. Most of the results 
are means of twenty experiments ; and several of them having 
been repeated at long intervals and by different hands, with- 
out any important difference in the means, the author is of 
opinion that they express the truth, at the same time that 
they may be open to some numerical rectification. 
