RESTITUTION-BODIES AND FREE TISSUE-CULTURE IN SYCON 317 
larva of the medusa Gonionema becomes transformed into 
a syncytial, undifferentiated mass, as was shown by Perkins (18). 
The obverse of this condition is shown by the failure of 
highly differentiated parts of the organism to maintain them- 
selves in the restitution-masses. Wilson and de Morgan and 
Drew found that portions of tentacles fail to become incorporated 
in the masses. This is paralleled by the failure of Hydra 
tentacles to regenerate. Apparently, on the one hand they 
are too highly specialized to dedifferentiate ; and on the other 
cannot exist as such in the conditions afforded by the restitu- 
tion-bodies. The nematocysts also are gradually resorbed in 
the restitution-bodies. 
If we seek to embrace the phenomena in one general view, 
we may say that Hydroid tissues in unfavourable or abnormal 
conditions lose much of their differentiation, come to have 
a low metabolic rate (in the general sense in which the term 
is used by Child (8)), and are more resistant. In these con- 
ditions specialized organs cannot exist. The same tissues 
in optimum conditions possess a higher metabolic rate, and 
are capable of maintaiming specialized organs such as the 
tentacles in existence. 
(d) “Normal’ and ‘Abnormal’ Phenomena. 
Attention has already been drawn to the fact that many of 
the processes occurring in restitution-bodies and free tissue- 
cultures run parallel with various phenomena of development. 
The normal phenomena constitute an interlocking series, each 
stage of which is determined by the preceding and helps to 
determine that which comes after. By studying processes 
which occur in * abnormal’ conditions, e.g. by dissociation 
methods, we remove the tissues of the organism from this 
developmental chain, where it is often impossible to say what 
occurrences are palingenetic, what adaptive, what the direct 
consequence of changes in the environment, and what con- 
ditioned by previous processes in the series ; by varying the 
conditions, we may then throw light upon the normal processes. 
So far my work has been mainly devoted to elucidating 
