96 
PACIFIC SCIENCE, Vol. VII, January, 1953 
tunately the type not only lacks the small 
chela, but the chitin of the anterior thoracic 
regions is so soft and distorted that the true 
nature of the orbital hoods could not be as- 
certained. 
Careful comparison of the specimen to the 
original description and plate of Edmondson 
showed only two minor differences: first, 
there was a faint longitudinal groove on the 
palm of the large chela that was not previous- 
ly noticed; second, the inconsistency of the 
specimen as to the armature of the merus of 
the third legs was not remarked upon. On this 
last point the omission is probably of no great 
moment, for in other species it has been ob- 
served that occasional specimens may lack a 
tooth on one or more legs, as this one does, 
yet the majority of the specimens will have it. 
Of the Hawaiian species this stands mid- 
way between C. paragracilis (Coutiere) and C. 
nanus Banner. It is like both species in the 
form of the rostrum and orbital hoods (prob- 
ably), in the mouthparts, in the general form 
of the antennular and antennal bases, in the 
general form (but not the specific sculptur- 
ing) of the large chela, in the general form of 
the third legs, especially in the curved pro- 
podus and the shape of the biunguiculate dac- 
tylus. It is like C. paragracilis in the length of 
the scaphocerite and in the tooth on the merus 
of the third legs, although in this species it is 
more poorly developed. It is like C nanus in 
the heavy form of the basicerite although not 
as heavy, in the breadth of the articles of the 
carpus of the second legs, and in the lack of 
the deep sculpturing of the large cheliped. 
However, it can be easily distinguished 
from these two species, as the lower concavity 
of the chela of C. hawaiiensis is slight and 
gradual, whereas in C. paragracilis it is delim- 
ited posteriorly by a lobe, and in C. nanus the 
entire depression is lacking; the second article 
of the carpus of the second legs is broader 
than long instead of slightly or markedly 
longer than broad, as it is in the other two 
species. From C. paragracilis it may be dis- 
tinguished also by the lack of spines on the 
inferior internal margin of the merus, by the 
shorter spine of the stylocerite, and by the 
ratio of the articles of the carpus of the second 
legs. It may also be distinguished from C. 
nanus by the shorter stylocerite and the longer 
scaphocerite. 
Schmitt (1939) has described a new subspe- 
cies, C. hawaiiensis clippertoni from Clipperton 
Island in the eastern Pacific. It was distin- 
guished from the type by two characteristics; 
the supposed lack on the type of longitudinal 
grooves on the palm and by the rostrum 
which in Schmitt’s specimen did not reach 
beyond the orbital hoods. The first of these 
differences is already removed, and it is likely 
that, if this species shows variation similar to 
that of C. paragracilis, the other difference will 
be within the range of variation for the spe- 
cies. However, until such a time as more speci- 
mens are examined, it would be best to leave 
the subspecies as it now stands. 
DISTRIBUTION: The type locality for this 
species is Lisianski Island, in shallow water. 
Crangon paragracilis (Coutiere) 
Fig. 33 a-h 
Alpheus paragracilis Coutiere, Soc. Ent. de 
Erance, Bui. (7): 149-152, 1898a. 
Alpheus paragracilis Coutiere, Fauna and Geog. 
Maid, and Laccad. 2(4): 883, pi. 76, fig. 22, 
1905. [Description same as original; figures 
added.] 
DESCRIPTION: Carapace produced into short 
acute rostrum almost reaching end of first ar- 
ticle of antennular peduncle.- Rostrum and 
carapace rounded dorsally, without carina. 
Shallow, rounded depression between base of 
rostrum and orbital hoods. Orbital hoods 
rounded, without processes. 
Stylocerite acute, extending to middle of 
second article of antennular peduncle. Pedun- 
cle proper short and heavy, with second arti- 
cle shorter than visible portion of first or third. 
Lateral spine of scaphocerite not over 0.2 
longer than squamous portion, as long as car- 
pocerite, definitely longer than antennular pe- 
