136 
ment. In the males a comparable but greater 
differential growth occurs, so that the fully de- 
veloped male chela is much heavier than that 
of the female, with a broader merus and palm, 
a shorter carpus, and with much heavier teeth 
on the fingers. In A. dorsalis it was found that 
chelae of the smaller males approached the 
form of the more mature females, and, in one 
case, the chela of an extremely large female 
was quite similar to the normal large males 
(Fig. 5e). This variation was not found to be 
perfectly correlated with size or maturity; for 
example, females with a definitely "immature” 
type of chela were found to be ovigerous, and 
specimens of the same size from the same lo- 
cality were found to be of varying development. 
On all species represented with a sufficient num- 
ber of specimens this variation was observed, 
with only A. djihoutensis showing it to a more 
minor extent. 
A variation somewhat similar to this was 
noted by Kemp (1915: 289 et seq.) for the 
males of his species A. polymorphus. He had a 
limited number of specimens and came to the 
conclusion that breeding activity, not growth, 
caused the polymorphism. We believe that the 
evidence given in the following pages indicates 
that the "polymorphism” must stem primarily 
from growth. Kemp was also unaware that his 
variation could be found in other species but he 
cautioned ". . . it is not unlikely that different 
forms of the same species have been described 
under separate names.” 
More perplexing is the apparent variation in 
the region of the anterior carapace. In the clas- 
sification of the species the breadth and length 
of the rostrum, and the development of supra-, 
extra- and infracorneal teeth, have been regarded 
as quite constant and used for specific distinc- 
tions. However, in a group of 30 specimens of 
A. dorsalis the tip of the rostrum was found to 
reach from the end of the first antennular article 
to the middle of the third antennular article; in 
the same group of specimens the extracorneal 
teeth were usually well developed and acute, but 
in one they were lacking. In A. areteformis some 
specimens carried an infracorneal tooth of vary- 
ing development, some did not (in three speci- 
mens from the same location two are without 
PACIFIC SCIENCE, Vol. XIV, April I960 
teeth, the third with a small infracorneal tooth 
bilaterally developed ) . 
To the contrary A. marshallensis, A. rhothi- 
onastes, and A. djihoutensis all present in some 
numbers in the collections, are quite uniform in 
the development of these parts. It would appear 
that some species are inherently more variable 
than others, a condition found in members of 
the genus Alpheus. 
Finally, in A. dorsalis the development of the 
tooth on the merus of the third legs and the 
proportions of articles of this leg were also 
found to vary. Like species in other genera of 
the family, the telson also varies in its pro- 
portions. 
These variations will be discussed more fully 
for each species in the following section. 
These variations in parts previously consid- 
ered to be of taxonomic importance render it 
difficult to decide upon a valid basis for species 
differentiation. A possible clue to aid system- 
atises may be found in the ecology of the forms 
being considered. At least in two pairs of closely 
related species the most valuable clue is their 
environment. A. dorsalis appears to be usually, 
if not always, associated with sea urchins of the 
genus Heterocentrotus, while the closely related 
A. indicus with the sea urchins of the genus 
Echinometra; again in the closely related pair, 
A. marshallensis and A. rhothionastes, A. mar- 
shallensis appears to be found only in the silty 
sandy shorelines of lagoons, associated with 
Alpheus strenuus Dana (whether it is a sym- 
biotic association or a mere ecological associa- 
tion is not known), while A. rhothionastes has 
been found almost exclusively on the coralline 
ridge of seaward reefs where it is subjected to 
heavy surf. This type of aid, however, is also 
questionable, for A. djihoutensis is widespread 
in its environments; most were collected from 
either the inner zone of outer reefs on islands, 
especially on the lee side of an island, or from 
somewhat similar conditions on the lagoon 
beaches of an atoll; some, however, were col- 
lected from near the surf line. 
With so much variation, with the extent of 
this variation apparently inconstant from species 
to species, and with so many species described 
on the basis of one or several individuals, it is 
impossible to decide which species put into the 
