6 
but it has as a rule a somewhat irregular contour. In Fig. 12 the 
distal segment is new; another similar case is shown in Fig. 13. In 
the latter the outline of the segment is more irregular, and in both 
the new segment is a little shorter than the one removed, and some- 
what thicker in its outer part. In Fig. 14 both the distal segment 
and a part of the next one are new; as indicated by the dotted line 
in the figure. 
In one case the leg had been cut off between the second and third 
segments (from the distal end). The leg was not thrown off, but the 
new part as shown in Fig. 15 is somewhat abnormal. There is a long 
distal segment representing, to all appearances, the distal segment of 
the normal limb. At its base there is a projection on each side, and 
on each projection a small terminal segment. The second segment 
has not been regenerated, and the abnormal condition at the base of 
the distal segment seems to be connected in some way with the absence 
of this segment. 
If the terminal segment of the first chelate appendage (the mov- 
able segment of the claw) be cut off the result is the same as for the 
last segment of the walking legs. The entire leg is not thrown off, 
and, the lost part does not reappear until the next molt. The other 
piece of the claw, the drawn-out end of the second segment may also 
be removed without the claw being thrown off. It too is not replaced 
until the next molt. 
If a leg from which the terminal segment has been cut off be 
examined several weeks after the operation the cut-end is found to 
be covered by a cuticle that bulges out in some cases. If the leg be 
decalcified and split in two with a sharp razor a small rudimentary 
leg will be found beneath the bulging cuticle. The interior of the 
segment is filled by a sort of whitish substance, the old muscles having 
broken down. It is this rudimentary limb that grows quickly at the 
next molt to replace the lost part. The irregular outline of the new 
segment is probably due to the confined space in which the rudimen- 
tary leg has developed. 
Conclusions. 
The results of these experiments show, it seems to me, in the 
clearest way that, for the hermit-crab at least, there is no relation 
between the regeneration of the leg and its liability to injury. Under 
the ordinary conditions of life it is almost impossible that a leg should 
ever be broken off inside of the breaking-joint. Nevertheless regener- 
ation of the limb will take place if the leg is cut off inside of this 
