52 
different lots of mice reacted differently. Thus with some mice 
0.001 gm. each of several thyroid preparations produced a maximum 
effect; in such a case it was necessary to repeat the experiments 
using 0.0005 gm. for example. In other cases the physiological 
effects produced by the feeding of 0.001 gm. of different thyroid 
preparations were very slight; in such cases the experiments were 
repeated with 0.002 or more gram. When the experiments were 
performed in this manner we found that the amount of acetonitrile 
from which the mice recovered under the influence of the thyroid 
was almost directly proportional to the amount of iodine the thy- 
roid contained; hence we conclude that the physiological activity of 
thyroid is proportional to the iodine content. The correctness of 
this conclusion was often confirmed as follows: If it required twice 
as much acetonitrile to cause the death of mice to which thyroid A 
had been fed as of those to which an equal amount of thyroid B had 
been fed, we found that the fatal dose of the nitrile was about the 
same in each case when two parts of B on the one hand and one 
part of A on the other were fed.® Results of greater scientific 
accuracy could probably have been obtained by a slightly different 
method of experimenting. Thus instead of comparing the total 
amount of thyroid which produced a given degree of protection, it 
would probably have been more accurate to first subtract the 
“maximum ineffective” dose (i. e., the maximum amount of thy- 
roid which could be fed without producing any demonstrable physi- 
ological effect) 6 and comparing only the amounts in excess of 
this dose. Similarly it would have probably been more accurate 
to subtract from the maximum nonlethal doses of the nitrile the 
maximum doses from which the mice would have recovered had 
they received no thyroid. In this way the amounts of the poison 
actually neutralized under the influence of the thyroid would have 
been compared. Such refinements of experimentation were, how- 
ever, not necessary for our purpose, which was simply to determine 
if there is a parallelism between the iodine content and the physi- 
ological activity of the thyroid 
It may be well to emphasize again some of the precautions neces- 
sary in this work. One of the most important of these is to have 
mice which have been kept under uniform conditions; the suscepti- 
bility of different lots of mice may vary widely even at the same 
season. Hence it is necessary to have a number of controls for each 
series of experiments. It is also necessary to work with doses 
which are not too near either the minimum or the maximum. 
® This method of double control is similar to the one adopted by one of us in com- 
paring the physiological activity of suprarenal preparations. (J. Am. M. Ass., 1906, 
47, p. 790.) 
b cf. J. T. Cash, Brit. M. J., Lond., 1908, i, p. 1213. 
