248 Proceedings of the Royal Society of Edinburgh. [Sess. 
limb, and when the injury is distal to the breaking-plane, suggesting that 
it does not occur when, for example, the diminished legs are cut, because it 
could serve no useful purpose. I find that in all the Pagurids which I 
have examined a definite plucking of a damaged fourth leg occurs after 
injury, and in this limb there is no breaking-plane. In addition, I find 
that in every case in which the eyestalks have been cut through, the 
chelipeds rise and tear at the seat of injury. At a later point “plucking” 
is discussed in relation to other acts of autotomy, and I hope to be able 
to show that in the Pagurid it is a return to an ancestral reaction type, 
which occurs when abnormal conditions arise. 
One other observation on autotomy in the hermit crab is of great 
importance. It was found that if the extensor of the second limb-segment 
of a cheliped be cut (this destroys the unisegmental mechanism which 
weakens the breaking-plane), and if the limb be damaged lower down, 
the stump is gradually picked away by the other sound cheliped. The 
process usually takes about a day to perform, but when it is complete 
the stump has been nicely dressed off down to the level of the breaking- 
plane. A large percentage of crabs die from haemorrhage before the 
process is complete, but in those in which it is carried out successfully 
regeneration takes place in the same manner as it does after autotomy. 
This “Autophagie” (Przibram) is a reaction of still lower type, and it 
occurs when all the mechanisms of autotomy have been disorganised. 
The above description of autotomy in the hermit crab shows the pro- 
found influence of environment on the animal. From a highly specialised 
type of reflex which is unisegmental, a plurisegmental one involving 
plucking movements can be produced by removing the animal from its 
shell. Autophagy after mutilation is a process which takes some time 
to perform, but it is the next reaction of the animal when all mechanisms 
of autotomy have been disorganised. 
Subsection 2. Galatheids. 
It is somewhat unfortunate that the majority of Galatheids found in 
the Clyde are too small for the anatomical study of the mechanisms of 
autotomy and for the analysis of autotomy reflexes by ordinary physio- 
logical methods, because in this group limbs seem to be abandoned on the 
slightest provocation and with the least difficulty. A consideration of 
certain phenomena to be seen in them leads to the form found in the 
Brachyura or “true” crabs. On this account it must now be studied. 
The Galatheids are lobster-like animals, but the abdomen is turned in 
loosely on the ventral part of the thorax, and by its rapid flapping move- 
