OXYGEN REQUIREMENTS OF SHEEEFISH. 
21 1 
and its lid put completely on, excluding bubbles. The entire apparatus was then kept 
in a bath of sea water whose temperature was maintained approximately the same as the 
water within the desiccator. The oxygen in the sample was meantime measured by 
titration. 
After a period, usually about an hour in length, the desiccator was two or three times 
inverted to render its contents uniform and to stimulate the shellfish to close and thus 
stop using oxygen. The exact time of the first rough movement of the apparatus was 
recorded as the closing time of the experiment. The outside of the desiccator was then 
dried with a towel and the entire apparatus weighed. When currents in the water had 
come practically to rest, the side stopcock was again connected to the same sampling 
bottle used at the beginning of the experiment and a sample taken as before. Since 
the capacity of the desiccators was sufficient for 1,200 grams of water, even with the 
large shellfish, and as the sampling bottles held approximately 300 grams, it is plain 
that water could run through the sampling bottle some time without emptying the 
desiccator. The stream indeed was allowed to run at least two minutes. The capacity 
of the desiccators was not enough to permit of taking a duplicate sample that would be 
reliable, but this was not necessary because there is little chance of error in the Winkler’s 
titration. 
The last step in the process was to empty the desiccator completely, dry it and the 
shellfish, and weigh them together. This gave a means of determining by difference the 
weight of the water. It was probably accurate to within 2 grams. As the dissolved 
oxygen of the water both at the beginning and the end of the experiment was calculated 
in parts per million, it was only necessary to multiply the weight of water used by the 
decrease in its oxygen content expressed in parts per million and divide the result by 
100 to find the decimilligrams of oxygen used by the shellfish. Experiment showed that 
no correction for change in the oxygen content of the water due to microorgarusms or 
physical factors was, under the circumstances of constant temperature, etc., necessary. 
Results were expressed in decimilligrams of oxygen per hour and in most cases also com- 
puted in decimilligrams per hour per 100 grams of shellfish and in some cases in decimil- 
ligrams per hour per i gram of the dried weight of the total shell contents of the organism. 
A formula for the first of these three computations might be expressed, therefore, as 
follows : 
100 i 
where A is the decimilligrams of O used per hour, W the weight, in grams, of water in the 
desiccator, o' the parts per million of oxygen in the first sample and a" in the second 
sample, and t is the time of the experiment in minutes. 
Throughout the experiments it was found necessary to observe the animal at fre- 
quent intervals, and if the shells closed up to discard the experiment, or if the shells 
closed only after a eonsiderable number of minutes had elapsed to terminate the experi- 
ment immediately. This precaution was necessary because, as is well known, certain 
of the shellfish used in this work can close water tight and in that condition take from the 
