Alpheid Shrimp, V— BANNER and BANNER 
rostrums: while its tip reaches to near the end 
of the second antennular article, well within 
the range specified above, it was considerably 
broader and thicker. Possibly more important, 
however, is the fact that the stylocerite also 
reaches only to the end of the second antennular 
article, instead of to the end of the third, a con- 
dition found in no other specimens. It may well 
be that these represent a new species or geo- 
graphic or ecological subspecies, but with only 
three specimens, all in very bad condition, it is 
impossible to decide with certainty. 
The other two nominal species also fall well 
within the range of variation. A. intermedins 
was separated by Yu from A. indicus by a 
slightly longer rostrum, reaching to one-quarter 
of the length of the third antennular article, and 
by the presence of 12 spinules on the propodus 
of the third leg. As shown in the listing above, 
both of these characteristics are very variable, 
and the separation of A. intermedins seems 
without basis. 
There is more doubt about the combination 
of A. kominatoensis with this species. Unfortu- 
nately Kubo did not contrast his species to 
Arete, where it belonged under the old classifica- 
tion (it has only four articles in the carpus of 
the second leg) but to Athanas. By contrasting 
the description and figures of Kubo with A. in- 
dicus the three differences are found: 1. The 
antennular peduncle is 2.2 times as long as broad 
instead of 3.0 (using the entire visible length 
of the peduncle divided by the breadth of the 
second article). 2. Kubo reports and shows a 
slight dorsal carina extending back to the middle 
of the carapace, while in A. indicus there is no 
true carina but rather a broad ridge terminating 
slightly behind the eyes. 3. The inferior margin 
of the merus of the third legs bears distally a 
slight tooth while in A. indictts it bears a strong 
tooth. It may be that these are valid criteria, 
especially the porportions of the antennular 
peduncle and dorsal carina — the development of 
the tooth on the third leg has been shown to be 
variable in the related A. dorsalis— and none of 
the specimens available closely approach the 
condition described by Kubo. However, in view 
of the wide variation already reported in this 
species and in A. dorsalis these seem to be ques- 
tionable characteristics for a specific separation, 
151 
yet it may be this form, being found as it is in 
the waters of southern Japan, is a separate sub- 
species. 
distribution : This species has been reported 
from Djibouti, Maldives and Laccadives, Indo- 
nesia, China, and southern Japan. In the present 
collection we have specimens from Canton, Ai- 
tutake, Bora Bora, and the northern Marshall 
Islands. They were found in heads of dead 
coral and associated with the boring sea urchin 
Echinometra. 
Athanas dorsalis (Stimpson) 
Figs. 5 , 6 
Arete dorsalis Stimpson, 1861. Proc. Acad. Nat. 
Sci. Philad. 12: 32. 
(?) Athanas mascarinicus Richters, 1880. Beitr. 
Meeresfauna der Insel Mauritius und der Sey- 
chelles Decapoda, p. 164. 
Arete dorsalis var. pacificus Coutiere, 1903. Bull. 
Soc. Philorn. Paris IX, 5(2) : 87, figs. 31-34. 
Arete maruteensis Coutiere, 1905^. Bull. Mus. 
Hist. Nat., Paris 11(1): 19, figs. 1-6. 
Arete maruteensis var. salibabuensis de Man, 
191CU. Tijdschr. Ned. Dierk. Ver. II, 11(4) : 
313; 1911, Siboga Exped. 39a 1 (2): 169, pi. 
4, fig. 13. 
DISCUSSION : To determine the validity of the 
separation of the three species and subspecies 
listed above, the specimens in the collection, 
especially a single collection of about 50 from 
Namu Island, Marshalls (Station 3167, J. E. P. 
Morrison), were examined closely. Of these 
specimens from a single locality, about 30 intact 
specimens were measured, and the findings con- 
firmed by examination of the specimens from 
other parts of the Pacific. 
Rostrum: In general the form of the rostrum 
is anteriorly a sharp triangle with an acute tip, 
while between the eyes the sides approach a 
parallel condition. In most the tip of the ros- 
trum reaches to the end of the first antennular 
article or to the first half of the second article; 
in three specimens it reaches well into the 
second half of this article, and in one it reaches 
the beginning of the third article. In most cases 
it is between 1.6 to 2.1 times as long as broad. 
Extracorneal Teeth: In most these are well 
developed, but to varying degree, and there is 
