DEDIFFERENTIATION IN PEROPHORA 651 
stages as follows: (1) shrinkage alone, (2) siphons closed, 
(3) siphons withdrawn from test, (4) spheroidal form assumed, 
(5) cessation of heart-beat, (6) reduction to stalked knob. 
It will be seen that this process is the reverse of that pre- 
viously described as stolon-resorption. In both eases, however, 
the equilibrium of the stolon-zooid system is altered, the altera- 
tion results in the resorption of one or other of its members, 
and this resorption may be total. 
Resorption of an organ like the stclon cannot be considered 
a very unusual phenomenon. It is paralleled, for instance, 
by the resorption of various larval organs at metamorphosis, 
such as the gills and tail of a frog-tadpole. Resorption of whole 
individual organisms, however, is much more unusual. So 
far as I am aware, it has only been noted at all adequately by 
Loeb (1900), who found it to occur in the Calyptoblast Hydroid 
Campanularia. I have re-investigated the phenomenon in 
Campanularia and also in Obelia, and can confirm the facts 
entirely. Something rather similar occurs in those Echino- 
derms where almost the whole of the larva is absorbed into 
the growing rudiment of the adult, but there remains an 
essential difference, namely, that resorption in such a case is 
determined as part of a normal development, whereas in 
Perophora and Campanularia it does not occur except as the 
result of circumstances which must be called abnormal. This 
is also true for the interesting observation made by Child 
(1904), who found in the chain-forming Turbellarian Steno- 
stomum that, if a cut be made through one of the zooids, 
the posterior half of such a zooid is completely resorbed by 
the zooid behind it. Resorption of whole zooids is also recorded 
(see later, p. 675). The case of Perophora is more remarkable 
than any yet recorded, partly owing to zooids being resorbed 
by subordinate systems, and partly owing to the great com- 
plexity of the zooids, which is very much greater than in 
Hydroids or Turbellaria. 
In all three cases, however—Ascidian, Flatworm, and 
Hydroid alike—the mechanism of resorption appears to be 
the same, namely, that the organs all decrease in bulk by the 
