1308 Wisconsin Academy of Sciences, Arts, and Letters. 
the determinations were volumetric, but- later, and principally, 
gravimetric methods were used. This was done by weighing the 
bottle empty and then adding a definite weight of water. After 
sterilization the bottles were again weighed and the loss deter¬ 
mined to within five cne-hundredths of a gram. In most of the 
experiments eight ounce Blake bottles were used. These were 
filled with a hundred grams of water. The results obtained in 
a considerable number of experiments are shown in the follow¬ 
ing table: 
Table a. 
Loss of Weight in Autoclaving 100 Grams of Tap Water. 
No. bottles. 
Greatest 
loss. 
Smallest 
loss. 
Average 
loss. 
20. 
7 grams 
7 
5 
5 
8 
5 
4.1 
6 
8.8 
4 grams 
1 
3 
4 
7 
3 
3.6 
4.4 
7 
5.20 grams 
4.65 
3.75 
4.00 
7.70 
4.00 
3.9 
4.80 
7.62 
20. 
10. 
10. 
10. 
10. 
3. 
10. 
10. 
Average total loss. 
5.07 or 5.07% 
It will thus be seen that the amount of loss is considerable and 
variable. An explanation of these facts is sought in the follow¬ 
ing experiments. 
II. A Comparison of Several Different Autoclaves. 
The autoclave used in the above experiments was a large (No. 
6) Kny-Scheerer Co. Steam Pressure Dressing Sterilizer (A). 
To show that loss noted was not due to any imperfections of the 
particular autoclave or the method of using it, several other 
sterilizers were tested, and in most cases these other autoclaves 
were run by the person ordinarily using them, and not by the 
writer. These other autoclaves may be described as follows: 
(B) A small size of the same type as above. 
YC) Similar in its general construction to the above but with¬ 
out a name plate. The steam in this case was generated 
by an electric generator. 
(D) A vertical cylinder after the French type. The steam is 
generated from a layer of water in the bottom. 
