Errect oF ARTIFICIAL MEMBRANE FORMATION Sl 
minutes after membrane formation behaved in a similar manner. 
Others, transferred eighty-two minutes after membrane forma- 
tion, gave few larvae. 
These experiments were repeated quite often, but the results 
were not always equally good. It is possible that the eggs of 
various females did not react equally well to this method. 
In the more recent experiments the eggs were first washed 
two or three times with ordinary sea-water after being taken 
from that containing potassium cyanide. 
In all these experiments the perfect regularity of the seg- 
mentation was very striking; the segmentation is almost as 
good as in eggs fertilized with sperm. This is an important fact, 
in view of the experiments in which the disintegration of the 
eggs after membrane formation is prevented by hypertonic 
sea-water. In this case the first division is often somewhat 
irregular. The experiments with potassium cyanide show 
that this irregularity is essentially a secondary effect of the 
hypertonic salt solution, and that membrane formation alone 
leads to a normal segmentation. 
Secondly it must be remarked that in these experiments the 
larvae stayed on the surface of the sea-water as in normal devel- 
opment, and gave large numbers of plutei. The shorter the 
time that the eggs had remained in the potassium cyanide 
solution, the larger the number of eggs that developed into 
normal plutei. The reason is obvious. Membrane formation 
sets going (or accelerates) the chemical reactions which deter- 
mine development. If we interrupt the oxidations in such an 
egg, there will accumulate in it decomposition products which 
ought to be removed by oxidation. I suppose that this accu- 
mulation of decomposition products leads to the injury of the 
egg, which is shown by the fact that the longer the eggs remain 
in the cyanide solution, the more their vitality is impaired. 
We have had the same experience with sea-urchin eggs ferti- 
lized with sperm, which also lose their vitality more quickly 
