DEDIFFERENTIATION IN PEHROPHORA 689 
germ-cells and interstitial cells disappeared, leaving only 
Sertoli cells and connective tissue. In one other case the 
germ-cells and Sertoli cells were much less affected than the 
interstitial tissue. This recalls the varying behaviour of 
the stolon-zooid system in Perophora according to the internal 
condition of the zooid. (See Lipschiitz, 1919, Chap. IV, where 
full references are given.) Another view of an almost identical 
problem is given by the varying response of the mammalian 
ovary to different intensities of X-ray treatment (Lipschiitz, 
1919, Chap. V, p. 205). 
The conclusions we reached in discussing Detwiler’s results 
(pp. 679-687) are of importance when we come to apply the 
principles of dominance, differential inhibition, and resorption to 
an explanation of the phenomena of metamorphosis. In meta- 
morphosis, as I have pointed out elsewhere (Huxley, 1921 d), 
we have to think of the full-grown larva as consisting of two 
minor systems in competition with each other—the differen- 
tiated system of larval organs, and the developing system of 
adult organs. The two enter into a state of balance. This 
balance may be tilted in favour of the adult, or kept at the 
existing tilt which favours the larval system. It has often 
been maintained that the time of metamorphosis was deter- 
mined by the production of a given relative quantity of some 
definite substance within the organism, e.g. thyroid secretion 
in the larvae of Amphibia. Such a concentration of a particular 
substance is often the effective agent in tilting the balance, but 
it is not the essential cause of metamorphosis. The essential 
cause of metamorphosis is that two mutually incompatible 
systems are in a state of dynamic physiological equilibrium 
within the same organism. 
In Echinoderm metamorphosis the mechanism for upsetting 
the balance appears to be simpler than in Amphibia. Experi- 
ments of Runnstrém (1917) and of my own, an account of which 
is now in the press, indicate that exposure of the pluteus tissues 
to unfavourable agencies of various descriptions will lead to 
their dedifferentiation and partial resorption. In nature the 
actual chain of events leading to this result appears to be as 
