CLIPPERTON ISLAND DECAPODS—CHACE 615 
females. Conversely, only four of the males in the collection are 
larger than the smallest certain female. But for these four males, 
whose carapaces measure 3.7, 4.0, 4.3, and 4.5 mm., the species 
would be considered protandrous. Of the 20 certain females, all - 
but two are ovigerous; those two have carapace lengths of 4.0 and 
4.7 mm. 
The rostral formula of these specimens varies as follows (the 
numbers in parentheses indicate the number of specimens with each 
ot (04), 262 ayy @); 27 @) 7? @); = ws 
arrangement): 
see (1); —— ae (1). The tooth in front of the hindmost one in 
Me Fa: series and one of the ventral teeth are apparently the 
last to appear. All but two of the 22 specimens with carapace lengths 
of 2.6 mm. or less have the second dorsal rostral tooth small or absent. 
Of the 35 larger specimens with intact rostra, only seven have the 
second dorsal tooth reduced or missing. These seven specimens 
range in carapace length from 2.7 to 4.5 mm.; only one of the four 
largest is a male, so the reduction of the second tooth is probably 
not a sexual character. All but three of the 22 specimens with less 
than two ventral rostral teeth are less than 3.0 mm. in carapace 
length, and only nine of the 28 specimens of this size have more than 
one ventral tooth. 
The branches of the upper antennular flagellum are fused to a 
variable extent in large specimens, but the length of the fused portion 
varies rather constantly with size in young individuals. In specimens 
with carapace lengths of 2.0 and 2.1 mm., the fused portion is longer 
than the free portion of the shorter branch; in one apparently aberrant 
specimen with a carapace length of 2.0 mm., the fused part is ten 
times as long as the free part but usually it is not nearly twice as 
long. At carapace lengths of 2.2 and 2.3 mm., the fused and free 
parts are approximately equal. From 2.4 to 3.3 mm. the fused part 
is always shorter than the free part but distinctly more than half as 
long. In larger specimens the fused portion varies from slightly 
more than a half to barely a third as long as the free part. 
The number of segments in the carpus of the second pereiopods 
is reasonably constant, and the variation that was observed is not 
correlated with size or sex. The number of carpal segments may be 
arranged as follows, in order of decreasing occurrence: 21, 20, 22, 
23, 24, 19. Of the 105 second pereiopods represented, no less than 
100 had the carpus composed of 20, 21, or 22 segments. One specimen 
had 19 segments on one side only, two specimens had 23 segments on 
one side only, and two specimens had 24 segments on one side only. 
Fewer than 20, and more than 22, carpal segments may therefore be 
