192 
PACIFIC SCIENCE, Vol. II, July, 1948 
passageway. Paul has demonstrated that upon 
severance of the appendage both the artery and 
nerve retract from the diaphragm, and blood 
extravasates from the ruptured distal end of 
the artery. The increased pressure in the haemo- 
coelic space forces together the funnel-like flaps 
which completely occlude the foramina. There¬ 
fore this mechanism supplants, in part at least, 
the earlier hypothesis (Reed, 1904) that a 
blood clot formed over the foramina, thus stop¬ 
ping the flow of blood. Actually some blood 
does pass to the outside of the diaphragm at 
autotomy because a layer of clotted blood covers 
the developing papilla for several days in P. 
crassipes, and has likewise been reported in 
other species. 
The subsequent papilla formation is accom¬ 
plished by the proliferation of cells from the 
free edges of the columnar epithelium. Emmel 
(1910) and Paul (1915^) have shown that 
blastematic cells which participate in regenera¬ 
tion of both the artery and nerve emanate 
from the epidermal layer, while the free ends 
of the severed artery and nerve contribute but 
little in the ensuing regeneration. The initial 
structure replaced is the diaphragm, a safety 
feature in the welfare of the crab; furthermore, 
differentiation within the papilla occurs from 
the fracture plane distally, never vice versa. 
Within a few days after autotomy, the epithe¬ 
lial cells form a layer over the stump and begin 
to proliferate at the center, initiating formation 
of the papilla, which is conspicuous a few days 
following autotomy. A significant feature de¬ 
scribed by both the above authors and sub¬ 
stantiated by the present investigation is the 
lack of sclerotization of the regenerating ap¬ 
pendage. This phenomenon is undoubtedly 
associated with the vast size increment from its 
papillary form to its functional manifestation 
after ecdysis. Throughout the entire papillary 
development the appendage is encased in a 
pliable sheath which is cast off during exuvia¬ 
tion. 
The initial information secured concerning 
regeneration in P. crassipes was derived from a 
study of 25 regenerating appendages on eight 
male and two female captive animals, each with 
one to five severed pereiopods. These animals, 
which were selected at random, ranged from 20.8 
to 30.4 millimeters in width. Their intermolt 
stage frequencies were: Q, one; early C 4 , three; 
late C 4 , three; D 2 , two; and D 3 , one. Several 
pereiopods were severed at ecdysis and thus pro¬ 
vided an opportunity for study of regenerative 
progress from stage A x . Daily observations 
included a check on the intermolt stage com¬ 
bined with a measurement and description of 
developing papillae. When observations were 
concluded 60 days later, five of the crabs were 
still living, three had succumbed during ecdysis, 
and two had died of other causes. Two crabs 
had molted and successfully regenerated their 
appendages, and several others were in advanced 
intermolt stages. Although the study was inter¬ 
rupted before adequate data concerning the 
relationships between regeneration and the 
intermolt cycle were secured, sufficient informa¬ 
tion was collected to clarify, in part, some of 
the nebulous data accumulated by pioneer in¬ 
vestigators who were unaware of the uncom¬ 
promising association between endysis and re- 
generational development. A description of the 
morphological features of regeneration in P. 
crassipes will be set forth first, followed by a 
synoptic account of its association with integu- 
mental development. 
Three days subsequent to autotomy the blood 
clot appeared completely black. The time inter¬ 
val varied, however, from 2 to 4 days, and 
coincided with the latter part of stage A 2 or 
the early part of stage B x . This clot was pushed 
outward from below by the proliferating epi¬ 
thelial cells until it was fractured on about the 
eighth day (but again the period varied from 
6 to 10 days). The papilla was conspicuous 
2 or 3 days later. The intermolt stage in which 
the papilla fractured the clot varied from B 4 
to Q. The interval between autopasy and the 
growth of the papilla to a length of 1 milli¬ 
meter required about 11 to 16 days, although 
one individual developed to this degree in but 
9 days. 
