83 
1914-15.] Kegeneration of the Legs of Decapods. 
after autotomy has occurred. Regeneration after autotomy at the breaking 
plane is also the process as it occurs in nature, though, as Morgan pointed 
out, new growth can take place from any point if certain precautions be 
taken to prevent removal of the limb at the breaking plane, and if life 
be preserved in the animal. 
Most decapods have at the base of their more vulnerable limbs this 
breaking joint, and here a membrane composed of two flaps stretches 
across the limb cavity. Through a foramen in this structure the nerve and 
vessel pass, and when the limb is autotomised these retract. Most writers 
say that a blood-clot then forms in the foramen and haemorrhage thus 
ceases (14), but my observations have led me to believe that in the majority 
of decapods, at least, such is not the case. In the hermit crab, for example, 
a beautifully arranged valve mechanism exists, and this, coming into action 
whenever autotomy has occurred, prevents the loss of even one drop of 
blood. Tait (15) in a recent paper failed to find any relation between the 
ease and frequency of autotomy in decapod crustaceans and the degree of 
development of clot-formation in the blood ; it may turn out, however, that 
there is a relationship quite different from what he imagined, and that the 
valve-mechanism now to be described is developed inversely as the clotting 
power of the blood. 
The epidermis of decapods consists of a single layer of columnar cells. 
On the outer side this secretes chitin and lime-salts, and so the shell is 
formed. At the breaking plane these outer hard layers are discontinuous 
in the greater part of the circumference of the leg, and the epidermis is 
modified. The bodies of the cells are greatly elongated and their pro- 
cesses seem to stretch inwards, joining with those of the opposite side 
(fig. 1). In short, they seem to be modified into connective-tissue fibres 
which mat together and form a diaphragm. Many writers on the subject 
have spoken of the epidermis being invaginated to form the diaphragm, 
and many have refrained from describing it altogether. From my own 
observations both in the developing and in the regenerating leg, I am led 
to the conclusion that the diaphragm is formed as described above, and 
corresponds with the “ collonades de soutien” described by Vitzou (16) in 
decapods. The following remarks on structure refer to Eupagurus, for in 
this species the mechanisms are most highly developed, and preparation of 
the tissue is most easy. 
The artery passes through a simple foramen in the diaphragm, which it 
completely fills, but not so the nerve. The latter has a funnel- like 
prolongation of the diaphragm fitting loosely round it, and passing back- 
wards and through this the venous blood from the limb must return, since 
