94 
PACIFIC SCIENCE, Vol. XVIII, January 1964 
times as long as broad, terminating in small but 
distinct tooth above dactylar articulation. Fin- 
gers not balaeniceps- shaped, bearing, however, 
slightly broadened opposing faces with slight 
row of setae on the margin; fingers equal in 
length to palm. Small chela of female is of 
similar form and armature, but about four-fifths 
as long as that of male. 
Carpus of second leg with ratio of secondary 
articles as 10:5.1:2.8:2.8:5.4. 
Third leg with small movable spine on is- 
chium. Merus unarmed, 3.7 times as long as 
broad, with distal end about 0.6 as broad as 
middle. Carpus 0.5 times as long as merus, 3.0 
times as long as broad, with relatively small 
teeth terminating superior and inferior margins. 
Propodus 0.7 as long as merus, about 6.0 times 
as long as broad, bearing 9 movable spines on 
inferior margin, scattered setae on superior 
margin. Dactylus 0.2 1 as long as merus, slightly 
curved, acute, and with slight thickening at point 
where secondary unguis might occur. Fourth 
legs similar to third. 
Telson 1.6 times as long as broad anteriorly, 
with posterior margin 0.6 times as broad as 
anterior; lateral margins slightly arcuate, with 
uniform curve; distal margin broadly arcuate; 
superior and distal spines usual. 
DISCUSSION: In the collection loaned to us 
from the Bishop Museum there was a group of 
five specimens from Wake Island, the type lo- 
cality for Dana’s species. Unfortunately, three 
of the specimens were small and the other two, 
which had been selected to be the neotype and 
allotype, were somewhat broken. However, the 
group of specimens from Johnston Island, which 
lies about the same latitude and only 1,600 miles 
away were almost identical with the Wake Island 
specimens and were in good condition. Both 
groups were used in preparing this description. 
Dana’s description of A. levius cuius is gen- 
eral enough to cover several species in the Ed- 
wardsii group. However, his figures are clear 
and do agree with our specimens except in two 
characteristics. First, there extends down the 
inner face of the large chela from the groove on 
the top margin a well-demarked U-shaped de- 
pression, while in Dana’s figure no depression is 
shown on the inner face of the chela. This 
difference is likely to be the result of faulty 
delineation. Second, the spine above the mov- 
able dactylus of the small chela is not shown. 
However, this character may be variable, as in 
A. paracrinitus Miers, or the chela as drawn by 
Dana could have been so rotated that the spine 
was not visible. 
Alp heus levius cuius Bate definitely does not 
belong to this species, being easily distinguished 
by the awl-like shape of the rostrum, the ex- 
tended flattened areas between the orbital hoods 
and rostral base, and the relatively smooth large 
chela. It should also be noted that A. levius cuius 
is found on the reef flat while Bate’s sole speci- 
not appear to be similar to any other species 
within the group or the genus, so we propose, 
therefore, to apply to it the name of Alpheus 
batesi. 
Our specimens also agree well with those of 
de Man (1911: fig. 98) except in the third leg. 
In de Man’s specimen the merus of the third 
leg was five times as long as broad, while the 
maximum length-breadth ratio found in these 
45 specimens is 4.4 times as long as broad. 
Alpheus levius cuius was considered by de 
Man to be closely related to A. has tar di Coutiere 
and A. bouvieri A. Milne Edwards. The points 
of difference between these nominal species in 
relation to the variations of the 45 specimens 
in this collection are discussed herewith: 
Rostrum length. The rostrum of A. levius cu- 
ius in Dana’s figures and in these specimens is 
roughly triangular and reaches only slightly 
beyond the middle of the first antennular article. 
The grooves beside the rostrum are shallow. 
The rostrum of A. bouvieri is reported to be 
similar. However for A. bastardi, Coutiere 
(1898) states that the rostrum is "Tres faible et 
tres obtuse, large, conique . . and that it is 
not separated from the orbital hoods by grooves, 
yet in the figures of both the 1898 and 1905 
works he indicated slight grooves. In these 
specimens the length of the rostrum, when 
compared to the first antennular article, appears 
to be influenced by the displacement of the 
carapace in capture (the carapace appears to be 
easily displaced, as it is in A. clippertoni [see 
under C. nanus Banner, 1953, fig. 30]). The 
grooves flanking the rostrum exhibit the ex- 
pected variation; some are very shallow, ap- 
proaching the condition that Coutiere indicated 
in fig. la (1898) and fig. 45 (1905) for A. 
