PACIFIC HERRING 
249 
It is obvious that the extreme variations shown in Figure 14 are significant, there 
being some underlying correlation between geographical distribution and number of 
vertebrae. The correlation is not an even one from south to north since the two 
localities farthest north, Golovin Bay and Eshamy Bay (Nos. 32 and 18, fig. 14) at 
o3 .F 3 
.9 & 
is S 
s ja 
> 8 
o « 
64° 30' and 60° 30 ' north latitude, have means of only 52.79 and 52.83, while Chignik, 
the Shumagin Islands, and Unalaska, far to the south, have means of 53.33, 54.67, 
and 53.22, respectively. Except for the Shumagin Islands and Golovin Bay all of 
the localities follow the same trend. The existence of a more or less systematic 
change with geographical location indicates the reality of the differences aside from 
any consideration of adequacy of sampling. 
