4 PROCEEDINGS OF THE NATIONAL MUSEUM VOL. 120 
(4) &, Sebastian Vizcaino Bay, lower Calif., coll. Jack Little- 
page, acc. no. 153 (AHF). 
(5) 2 #o’, 3 99, San Cristobal Bay, lower Calif., 41 fms (74 m), 
trawl, Apr. 27, 1950, Velero Sta. 1949-50 (AHF). 
(6) o, 2, 1 juvenile, 27°24’ N, 114°40’ W, San Cristobal Bay, 
lower Calif., 40 fms (72 m), Aug. 20, 1960 (1800-1820 hrs.), otter 
trawl, 6008B, 123.37, coll. Fred Berry (SIO). 
(7) &@, Turtle Bay, lower Calif., in purse seine inside bay dur- 
ing daytime, Stella Maris, coll. Ben Fukuzaki (AHF). 
(8) 9, 6.25 mi SSW San Hipolito Pt., 36 fms (65 m), trawl, 
Apr. 29, 1950, Velero Sta. 1952-50 (AHF). 
The following specimen, kindly loaned by Dr. John C. Yaldwyn, 
is the first recorded from New Zealand: 9, New Zealand waters, 
DM reg. no. Z.Cr.1493. 
Morphological Differences—Initial Study 
Only four possible distinguishing features were noted on preserved 
specimens from different areas. In each case there was considerable 
variability and such overlap between the series that statistical analyses 
were deemed necessary. 
Initially, comparisons were made between Australasian and Ameri- 
can material. Chilean and Californian collections were pooled for 
the following reasons: (1) the single Panamanian specimen tends to 
bridge the biogeographic gap; (2) probably the northern population 
extends in strength beyond the known southern limit as evidenced by 
the fact that 15 of the 37 ‘‘Californian” specimens have been collected 
in recent years in Mexican waters; and (3) the small number of 
Chilean specimens (9 only) available for study. 
In each case a quantitative study of a feature was made and group 
means were computed. In most cases deviations from means ex- 
hibited by each of the individuals in a group were plotted as frequency 
distribution histograms; in other cases raw data were plotted similarly. 
These histograms sometimes indicate differences in the constitution 
of populations from the two areas (figs. 1-3). Histograms showing 
normal distributions are not given. 
Variabilities of populations were computed in terms of standard 
deviations divided by means; these again indicate some population 
differences. 
Finally, ¢ tests of the significance of differences between group 
means were carried out (Lacey, 1953, p. 114; Fisher and Yates, 1957). 
It is appreciated that the applicability of this test is affected by ab- 
normal frequency distributions; however, the low p values obtained 
