NO. 3564 HEMISQUILLA ENSIGERA—STEPHENSON 5 
in all relevant cases (p<0.001) leave no doubt of the significance 
of the differences. 
In certain cases inspection of raw data indicated the possibility 
of sexual dimorphism. Means for Australasian males and females 
and for American males and females, respectively, were computed 
and ¢ tests carried out in “promising” cases within the locality 
groupings. 
In other cases it was suspected that the measured ratios depended 
upon specimen size. Using all specimens from a given area, cor- 
relation coefficients were computed between ratios and carapace 
length (i.e., the measure of specimen size) and were tested for sig- 
nificance (Fisher and Yates, 1957). Regression coefficients were 
computed in several cases. 
MANDIBULAR PALP.—Kemp (1913) noted two-segmented palps in 
an Australian specimen and from one to three segments in Chilean 
specimens. Schmitt (1940) noted three segments on the right and 
two on the left side of a single Southern Californian specimen. 
NUMBER OF INTERMEDIATE TELSON LOBES 
TTT 
ll AUSTRALASIAN OUT AUSTRALASIAN 
—————j AMERICAN B = | AMERICAN 
Ficure 1.—Frequency distribution histogram showing percentages of specimens with 
varying numbers of: A, segments of mandibular palp; B, intermediate telson lobes. 
Figure 14 shows that the Australasian population gives an approx- 
imately normal distribution curve with a mode of about two segments, 
but the American is strongly skewed to the right. Table 1 indicates 
approximately equal variability in the two populations. On the 
difference between means, ¢ tests give p<0.001. There is no doubt 
that the specimens belong to populations that are statistically separate. 
