504. G. H. PARKER 
carried to an extreme it temporarily reduces the colony to a 
mere fraction of its original volume. A stimulus that will bring 
about a withdrawal of zoéids will not always induce a contraction 
of the whole colony, so that these two processes must not be 
regarded as invariably concomitant. 
The passage of impulses from any part of the colony into the 
individual zodids is not interfered with by making cuts in the 
colony so long as the resulting pieces still retain organic connec- 
tions. Incisions may be made to any number or extent and the 
rachis may be cut into narrow strips or complicated forms, but, 
so long as organic continuity is retained, impulses will spread 
from the region of stimulation to the most distant autozodid 
and bring about its withdrawal. In short the spread of this form 
of impulse, so far as experimental pattern is concerned, is exactly 
like that of the spread of the impulses for phosphorescence. Even 
in preparations in which the colony is cut almost in two along 
its-chief axis and the halves remain attached only through the 
distal tip of the peduncle, the stimulation of one half-rachis 
involves not only the contraction of the autozodids of that half, 
but, after a brief interval, the contraction of those of the other 
half. This reaction suggests that the impulses to zodid contrac- 
tion run in waves as do those for phosphorescence, and this is 
probably true, but the zodids contract so slowly as compared 
with the flashing of the phosphorescent points that, under ordi- 
nary circumstances, the undulatory nature of the impulse is 
quite lost sight of. 
The impulse to zodid withdrawal is checked by magnesium 
sulphate in the same manner as is that for phosphorescence. If 
a rachis is cut in two except for a small bridge of tissue over which 
impulses to zodid contraction can be shown to pass and this 
bridge is covered with crystals of magnesium suphate, in approxi- © 
mately five minutes the impulses begin to fail to pass and in ten 
minutes no zoéid contractions occur on one side of the bridge 
when the other side is stimulated. After about half an hour in 
pure sea-water, transmission over the bridge of tissue is again 
resumed. The same kind of temporary block can be established 
at the unsplit distal end of the peduncle of a partly bisected colony 
