146 
PACIFIC SCIENCE, Vol. II, July, 1948 
free from their attachments; all parts were 
placed on blotting paper and allowed to drain 
for 5 minutes; the gills were then dropped into 
one graduate partially filled with water, and 
the remaining body parts into another graduate 
similarly filled; the volume was recorded as 
the difference in the column of water. The 
results are summarized in Table 1. 
Unfortunately, crabs more highly adapted to 
a semi-terrestrial existence than P. crassipes were 
not available for this study; consequently, infor¬ 
mation was secured from statements concerning 
gill reduction in the near-terrestrial types made 
by Pearse (1929 a). He has shown that some 
fiddler crabs (Uca minax, U. pugnax, and U. 
pugilator) have but 12 gills. Therefore, with 
the exception of P. crassipes, it may be assumed 
that in general crabs adapted to lengthy expo¬ 
sures to the air will have the gill number re¬ 
duced. It follows that a reduction in the num¬ 
ber and volume of gills is undoubtedly a critical 
factor in the selection of a habitat; i.e., H. nudus 
and H. oregonensis can withstand relatively long 
periods of desiccation because of a gill reduc¬ 
tion, and consequently are found in the "middle 
high” intertidal zone, where they are exposed 
for many hours each day. P. crassipes, on the 
other hand, presents an enigma in that the 
usual number of gills (18) for strictly aquatic 
forms is not altered; however, the total volume 
of the gills in relation to the total volume of 
the body is the least of any of the animals 
examined; and the anterior arthrobranch of the 
second thoracic segment, the gill which is lost 
in both species of Hemigrapsus, is minute. 
Apparently, evolution with regard to the gills 
in P. crassipes has tended toward the reduction 
in size of all gills in lieu of reduction in num¬ 
ber. The proportional percentage of gill volume 
to body volume in this species is somewhat less 
than in either species of Hemigrapsus, but seems 
inadequate to account for the tolerance of the 
former type to a drier environment. The slight 
difference between the gill-volume-body-volume 
relationship in the three grapsoid types corre¬ 
sponds with their similarity in resistance to 
desiccation. In fact, in the Puget Sound area 
where P. crassipes does not occur, H. nudus 
apparently ranges in the tidal horizon which 
would normally contain P. crassipes; however, 
the individuals of the former species at this 
higher elevation are smaller than those within 
the optimal environmental frontiers for the 
species (Ricketts and Calvin, 1939). 
Sex Ratio 
Sexual records from collected specimens were 
kept for 10 months of the year, not only to 
ascertain the relative numbers of the sexes, 
but to secure information on the supposed 
tendency of ovigerous females to sequester them¬ 
selves. Virtually all the specimens were col¬ 
lected at night with the aid of a flashlight; 
hence no sexual discrimination occurred while 
collecting. A total of 2,224 animals was col¬ 
lected, of which number 1,141 were males and 
1,085 were females. The nearly even sexual 
distribution (males exceed females by approxi¬ 
mately 3 per cent), together with the fact that 
the females equal or exceed the males during 3 
widely separated months, appears to indicate 
that monthly inequalities in sex ratio are insig¬ 
nificant. Furthermore, during August, when 
1,264 crabs were collected at random, the sex 
ratio was virtually even, indicating that the 
ovigerous females at the height of the breeding 
season (28 per cent of all females collected 
were ovigerous) do not exhibit sequestering 
habits; indeed, the proportionate number of 
females actually increased during the breeding 
season. 
INTERMOLT CYCLE 
The phenomenon of molting and filling of 
the haemocoelic spaces by imbibition of water 
has been considered by most workers to be a 
recurring event in the life of Crustacea with 
the intervening period between actual molts a 
more or less static one. However, the profound 
transformations occurring within the Crustacea 
from the conclusion of one molt to the termina¬ 
tion of the one following are a continuous series. 
