250 University of California Publications in Zoology  |Vou.9 
that end a glass cap, the suppression of the distal hydranth being 
coupled with an acceleration in the development of the proximal 
hydranth. The latter experiment is open to the possible objection 
that while preventing oxygen from diffusing inward, the glass cap 
might hinder the diffusion of gases from the stem outward. 
A recent experiment, however (performed during a laboratory 
course In experimental zoology with the codperation of several 
students) shows that this objection is not valid, and points con- 
clusively to the supply of oxygen as the essential factor in deter- 
mining the result. 
The method employed consisted in separating by a partition 
two bodies of sea water that should differ only in the absence of 
oxygen from one of them, and then passing segments of Tubularia 
stems through perforations in the partition so that approximately 
equal lengths should project on either side of it, their distal ends 
being immersed in the deoxygenated water. The simple device 
used to obtain these conditions was composed of a small battery jar 
and, fitting snugly into it, a erystallizing dish whose bottom had 
been removed and a sheet of soft paraffine about 5mm. thick sub- 
stituted for it. A portion of water boiled a few minutes to remove 
the contained oxygen completely filled the battery jar below the 
paraffine plate. Another portion of the same boiled water, but 
aerated by pouring from one dish to another, almost filled the 
crystallizing dish above the paraftined plate. The Tubularia stems 
were so placed that their distal ends projected downward into the 
de-aerated water, their proximal ends upward into the aerated 
water. Asa control, the previous conditions were duplicated with 
the difference that aerated water only was used. 
Accordingly, two sets of nine pieces each, approximately 2 em. 
long and equal in size and vigor, were arranged as described. 
Sixteen hours later the control pieces possessed six distal, but no 
proximal hydranths; the other pieces, on the contrary, possessed 
two proximal but no distal hydranths. Four hours later, seven 
distal hydranths and six proximal had appeared in the control; 
on the other pieces there were now five proximal, but still no 
sign of distal hydranths. Three hours later the control pieces 
were apparently in the same condition, while in the other dish a 
