248 
BULLETIN OF THE BUREAU OF FISHERIES 
VERTEBRAE 
RELIABILITY OF VERTEBRAL COUNT 
Of the four characters chosen to show structural differences, the vertebral count 
is the most reliable. The count is not altered by preservation and can be made with 
absolute accuracy and with greater ease than those of the fin rays. The number of 
vertebrae is determined at a very early stage, before or shortly after the hatching of 
the ova, so that it is not altered by any subsequent environmental conditions. Since 
the ova are attached, it is obvious that one adult population can not contribute, 
before the characters become fixed, to another adult population with different 
structural characters, as might be were the ova floating freely in the currents. On 
account of these advantages the differences in the vertebral count have received the 
chief emphasis in this analysis. 
There are now available for comparison 10,132 vertebral counts of the Pacific 
herring. Of these counts, those from San Diego Bay, Puget Sound, and Alaska, 
totaling 7,960, are original. Those from Monterey Bay, the first two San Francisco 
Bay counts, and those from the Straits of Georgia are by Carl L. Hubbs (1925); the 
last San Francisco Bay count and the remainder of the British Columbia counts are 
by William F. Thompson (1916). 
COMPARISON ALONG WHOLE LENGTH OF COAST 
The average number of vertebrse increases to the northward and westward from 
San Diego. For the purpose of showing this, Table 6 and Figure 14 are presented. 
Table 6 gives the frequency distributions of vertebral counts in various localities 
and Figure 14 gives the means for each locality plotted against the distance from 
San Diego, following the general trend of the coast. The trend of the line of means 
has been drawn in by inspection. 
Table 6. — Variation in number of vertebrse in all samples of all localities 
No. 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
29 
30 
31 
32 
Locality 
Vertebras 
Num- 
ber 
Mean 
Prob- 
able 
error 
45 
46 
47 
48 
49 
50 
51 
52 
53 
54 
55 
56 
57 
12 
148 
209 
38 
1 
408 
50.68 
0. 023 
2 
20 
41 
25 
1 
S9 
51.03 
.06 
1 
1 
4 
18 
256 
410 
125 
5 
820 
50.78 
.019 
1 
3 
25 
65 
6 
100 
51.71 
.052 
1 
1 
28 
354 
738 
138 
3 
1,263 
51.78 
.01 
7 
23 
17 
3 
50 
52. 32 
.075 
2 
22 
176 
126 
18 
344 
52.40 
.026 
13 
173 
239 
36 
2 
463 
52.66 
.021 
2 
20 
228 
381 
50 
1 
1 
683 
52.68 
.018 
9 
108 
210 
24 
1 
352 
52.72 
.023 
6 
15 
4 
25 
52.92 
.078 
1 
3 
3 
86 
459 
368 
40 
2 
962 
52.33 
.017 
2 
10 
8 
5 
25 
14, 43 
. 163 
17 
75 
74 
11 
177 
52. 45 
.038 
1 
13 
118 
220 
48 
1 
401 
2. 76 
.024 
1 
14 
116 
192 
42 
2 
367 
52. 72 
.027 
1 
1 
16 
139 
143 
99 
322 
52. 55 
.029 
1 
41 
91 
16 
1 
150 
52. 83 
.035 
5 
45 
143 
30 
1 
224 
52. 90 
.030 
3 
41 
75 
19 
138 
52. SO 
.040 
1 
7 
66 
57 
6 
137 
52.44 
.040 
1 
9 
34 
51 
5 
100 
52. 50 
.051 
1 
1 
26 
228 
378 
98 
8 
740 
52. 78 
.031 
1 
20 
180 
261 
66 
3 
531 
52. 72 
.022 
2 
21 
52 
12 
87 
52. 85 
.048 
5 
42 
82 
35 
1 
165 
52.91 
.041 
1 
30 
63 
17 
3 
1 
115 
52.95 
.050 
4 
5 
55 
39 
3 
1 
107 
53. 33 
.053 
1 
7 
36 
130 
213 
62 
7 
456 
54. 67 
.029 

1 
5 
2 
8 
53. 13 
20 
106 
54 
3 
183 
53.22 
.032 
2 
43 
77 
18 
140 
52.79 
.038 
Standard 
deviation 
of dis- 
tribution 
0. 691 
'*.'797 
.768 
"’.’786 
.712 
.681 
.683 
.634 
.580 
.769 
1.205 
.750 
.719 
.755 
.772 
. 636 
.657 
.693 
.692 
.763 
. 790 
.757 
.670 
.776 
.790 
.806 
.928 
'.’650 
.671 
