116 
PACIFIC SCIENCE, Vol. II, April, 1948 
this season, and many of them were in ad¬ 
vanced stages of maturing of the sexual products. 
The smallest fish, under about 650 mm., which 
are believed to be fish only a year old, had un¬ 
developed gonads and the determination of sex 
was difficult. 
Measurements were made in millimeters with 
the same slide-calipers used by Godsil and 
Byers (1944), or with dividers for short dis¬ 
tances such as diameter of iris or length of 
maxillary. The various measurements and 
counts were made in the exact manner described 
by these authors in their Appendix II except as 
noted below. 
Pectoral insertion to insertion first dorsal was 
measured with the tip of the fixed arm of the 
caliper at the insertion of the first dorsal fin, the 
sliding arm being brought to the anterior ter¬ 
mination of the dorsal margin of the pectoral 
fin. 
Length of longest (first) dorsal ray was 
measured with dividers from the juncture of the 
ray (with the fin extended) and the contour of 
the body to the tip of the ray. The longest ray 
was the first ray in all cases except two, which 
are indicated in the table. 
I refer to the ’'length” of the second dorsal 
and anal where Godsil and Byers use the term 
"height.” 
Length of longest dorsal finlet (the 5th or 
6th) was measured with dividers from the an¬ 
terior margin of the finlet to the tip of the pos¬ 
terior filament. The number of dorsal finlets 
and number of anal finlets were counted from 
posteriorly forward and if the last, most an¬ 
terior, finlet (the last two in some cases) was 
attached to the second dorsal (or anal fin) by 
a membrane, it is recorded separately from the 
count of the free finlets. Thus if there were 9 
finlets, all free, the record is 9; if there were 9 
free finlets and one attached by a membrane to 
the fin, the record is 9+1. This seems not to 
be a very good character because of the diffi¬ 
culty of distinguishing between attached finlets 
and the posterior end of the fin itself. It may be 
noted that our counts tend to average higher 
than those of Godsil and Byers, even when the 
difference in method of counting is taken into 
account. There also seems to be some difference 
between the author and Mr. Marr as to when a 
finlet is considered attached. 
The character "length of base of first dorsal” 
would be more accurately called "first dorsal in¬ 
sertion to second dorsal insertion” since, follow¬ 
ing Godsil and Byers, that is the measurement 
taken for this character. 
Weights were usually taken with a spring 
balance reading in pounds. Fish over 100 
pounds were weighed in pounds on a platform 
scale. Some small fish were weighed on a small 
spring balance, read to 0.1 kilo. Because the 
readings recorded in pounds to the nearest 
pound were later translated to kilos, the ac¬ 
curacy is slightly less than indicated; individ¬ 
ual weights may be in error by as much as 0.25 
kilo. 
RELATIVE GROWTH 
The measurement data in Godsil and Byers’ 
paper (1944) are recorded in terms of body 
proportions, that is, in terms of the times a 
given measurement is contained in the body 
length, or in the head length, depending on 
the character. Body proportions have also been 
used to characterize supposed species, for ex¬ 
ample, by Jordan and Evermann (1926) and by 
Nichols and La Monte (1941). Where data 
from fish of different sizes are compared, this 
is unsatisfactory unless the ratio of the dimen¬ 
sion under consideration bears a constant ratio 
to the dimension, such as total length, which is 
employed as a standard. Where such a constant 
ratio does not exist, it is necessary to compare 
only fish of the same size or, which is more effi¬ 
cient, to compare the regression of the given 
character on fish length (or head length). 
Nichols and La Monte have recognized this in 
the case of the length of the second dorsal and 
anal fins and have combined a fish size with a 
ratio of fin length to body or head length in 
drawing up species descriptions, and have made 
