602 
base of the fifth segment. A part of the chitnous covering of this 
segment still remains, but most or all of the contents of the segment 
seems to have disappeared, so that the new bud arises from or near 
the outer surface of the breaking-joint. 
Two further cases of regeneration of the larger of the first legs 
are shown in Figs. 9 and 10. In the first of these the leg had been 
cut off at the fourth segment, and from this level the new leg arises. 
In the other case the cut had been made through the middle of the 
fourth segment and a rather broad projection of new tissue marks the 
formation of the new distal portion of the leg. The four remaining cases 
in this experiment, shown in Figs. 11—14, represent the regeneration 
of new legs from the fourth and fifth segments. 
ES = ID 
2 u: 
aX a N ie 
9 10 i 72, 738 14 
These results taken in connection with those described in one of 
my earlier papers show beyond a doubt that regeneration of the leg 
of the hermit-crab can take place at a number of different regions 
lying outside of the breaking-joint and presumably, therefore, at all 
levels. The chief interest in these results lies in their bearing on the 
hypothesis of a supposed connection between regeneration and liability 
to injury. As previously pointed out the experiments indicate very 
clearly that there can be no such connection in the power of rege- 
neration possessed by different regions of the leg of the hermit-crab, 
for the legs can rarely or never be broken off outside of the breaking- 
joint under natural conditions without the leg being thrown off at the 
base. It is, of course, possible that the leg might be so injured that 
the muscles at the base are thrown out of action in which case the leg 
could not he cast off, but the chance of such an accident must be so ex- 
tremely remote that we can afford to leave this possibility out of 
account. Direct observation has shown, furthermore, that under natural 
conditions, the legs of the hermit crab are always thrown off at the 
breaking-joint. It is not desirable to discuss here these results in 
their wider bearing since I have treated this subject very fully in my 
forthcoming book on “Regeneration” and the special chapter in the 
TU 
