Biology of Pachygrapsus crassipes —Hi ATT 
195 
Fig. 17. Comparative growth of normal and regen¬ 
erating appendage of P. crassipes at ecdysis. A, growth 
of the third pair of pereiopods; B, growth of the 
chelae. The percentages indicate the size of the regen¬ 
erating limb compared with the opposite normal limb 
at a certain intermolt interval. L.P., left pereiopod; 
R.P., right pereiopod. 
was undertaken, sufficient observations on both 
captive and wild crabs have been made to de¬ 
scribe this feature in some detail. With the 
exception of the dactyls, regeneration from any 
point in the appendage other than the fracture 
plane has never been recorded in an examina¬ 
tion of well over 3,000 crabs in collections and 
of countless numbers observed in the field. It 
seems probable, therefore, that under normal 
conditions this species always severs the limbs 
at the fracture plane. The regeneration of por¬ 
tions of the dactyls is common in nature, and, 
because of their utility in diagnosing intermolt 
stages (p. 152), several crabs which had regen¬ 
erating dactyls were available for laboratory 
observation over extensive periods. The papilla 
of the dactyl grows approximately to its normal 
size; but, since the regenerating portion of the 
dactyl is invested by an external membrane 
similar to that covering the papilla at the 
fracture plane, the papilla displays no spines. 
The investing membrane is cast at ecdysis and 
the new dactyl emerges in its normal form. 
In wild crabs the frequency of regenerating 
dactyls is highest for the chelae, doubtless the 
result of the vigorous activity involving the 
chelae which often leads to the fracture of the 
distal portion of this podomere. One such 
example is shown in Plate 2, Figure 2. It is 
significant and distinctly advantageous to the 
crab that an injury to the dactyl does not result 
in autotilly or autotomy. It was found that 
injury to all podomeres except the dactyl stim¬ 
ulated virtually immediate amputation at the 
fracture plane. It has long been known that 
the thoracic nerve must be stimulated to effect 
both autotilly and autotomy; and since branches 
of this nerve do not extend into the dactylus 
(Pearson, 1908), injury to this distal podomere 
is insufficient to activate the mechanism of am¬ 
putation. 
Of the several thousand crabs handled during 
this study, approximately 30 per cent had one 
or more regenerating appendages. Such a high 
frequency of severed appendages implies a 
relatively precarious existence which is reflected 
in the ease of amputation and relatively fre¬ 
quent molts. If, for example, comparisons are 
made between P. crassipes and those brachyuran 
species residing lower on the strand—or below 
low-tide level—with respect to type of habitat, 
activity, predation, and capability of self-ampu¬ 
tation, marked differences are manifest imme¬ 
diately. Hundreds of crabs ( Cancer magister 
Dana) brought into wholesale markets by 
fishermen have been examined for regenerating 
appendages. Relatively few regenerating append¬ 
ages were found; indeed, to find one was rather 
rare. Autopasy is less easily accomplished in 
this cancroid species and much more time is 
required for this activity than for the process 
involved in P. crassipes. Thus the infrequent 
regeneration in C. magister is probably asso¬ 
ciated with the difficulty of self-amputation, 
which in itself may be an adaptation to the 
infrequent molts and the comparatively more 
placid existence of this crab. Other cancroid 
crabs ( C. productus Randall and C. antennarius 
Stimpson) which occur in the intertidal zone 
were examined similarly and found to exhibit 
characteristics comparable to C. magister. 
Differences between P. crassipes and the more 
sluggish and less-exposed cancroid crabs signifi¬ 
cantly portray many aspects of the evolutionary 
