196 
Fig. 3. Chum salmon branchiostegal, showing 
modified configuration of anterior rays. 
tical analyses of between-sex, within-sample, 
differences were not attempted due to small 
sample sizes; nor were be tween -area differ- 
ences tested for statistical significance, inasmuch 
as many samples were obtained from the high 
seas and contained specimens of unknown and 
perhaps differing origins. 
For each sample, probability values associated 
with the t-distribution for differences between 
paired counts were obtained by sex. Sexes were 
then combined and analysed as a single sample. 
Similarly, for each species, probability values 
were determined for all males and females sep- 
arately; the sexes were then combined to de- 
termine the mean differences between left and 
right side structures for comparison among 
species. 
RESULTS 
The results of the analyses of observed dif- 
ferences between left and right side structures 
are discussed for each meristic character. The 
occurrence, direction, and significance of the 
Fig. 4. Schematic presentations of aberrancies oc- 
curring in pectoral fins, a, "Floating” ray; b, fusion of 
lateral processes from adjacent fin ray bases. 
PACIFIC SCIENCE, Vol. XX, April 1966 
mean differences between left and right side 
counts have been examined by species, sample 
location, and sex. 
Bilateral Variation in Lateral Line Scales 
In sockeye salmon (Table 1), the percentages 
of specimens exhibiting asymmetry in lateral 
line scales and the direction of asymmetry 
varied among samples of each sex and among 
combined samples of both sexes from different 
locations. As shown by the right-hand column 
(headed 100 P), the probability values for the 
observed mean differences were not significant. 
For the combined samples, males with differ- 
ences in left and right side counts averaged 
69%, and females averaged 65%. Asymmetry 
between left and right side scale counts oc- 
curred in approximately 67% of the total of 
395 specimens. The direction of asymmetry In- 
dicated that the right side developed approxi- 
mately 0.12 scales more than the left side. This 
difference could have occurred by chance alone 
more than 5 out of 100 times (P greater than 5); 
thus, the left and right side variations in lateral 
line scale counts were not significantly different 
In sockeye salmon. 
Examination of data from chum salmon scale 
counts (Table 1), by sample location and sex, 
indicated some disparity among samples within 
these categories. Mean differences between left 
and right side counts of males were significant 
in one sample from the high seas. In this sample, 
collected by the MY "Paragon,” the mean dif- 
ference for females was not significant. In the 
Hokkaido Island sample both males and females 
