Vertebral Number of Nehu — Tester and Hiatt 
61 
Fig, 2. Vertebral column prepared by the "photographic” method. Note the anterior (a) and posterior (b) 
end points in counting. 
solving 1 gram of Alizarine Red S. in 500 
milliliters of 3 per cent potassium hydroxide 
solution. The staining solution consisted of 
10 milliliters of the stock solution added to 
100 milliliters of 3 per cent potassium hy- 
droxide. 
In counting the vertebrae, both end points 
were clearly defined. The first vertebra was 
readily identified by its high spine with a 
small triangle of cartilage both anterior and 
posterior to it. The last vertebra was identified 
as the one immediately anterior to the up- 
turned, segmented urostyle. These end points 
may be seen in Figure 2. Thus the count, 
which was checked at least once for each fish, 
included neither the basioccipital nor the uro- 
style. 
Both methods of preparation were satis- 
factory. However, the first method was time 
consuming, and in several cases it was diffi- 
cult to get the column in clear focus along 
its entire length for photographing. The 
second method is recommended. With its 
use, a small number of vertebral columns with 
abnormal centra (6 cases out of 851) were 
detected, and the counts on these abnormal 
specimens were discarded. 
Length Measurements 
Length measurements were made on all 
samples, but in one from Kaneohe Bay 
(10/25/47), they were not identified with 
vertebral number. Standard length is defined 
as the distance (measured to the nearest milli- 
meter) from the tip of the upper jaw to the 
posterior edge of the hypural plates. In speci- 
mens which were photographed, the length 
was determined with dividers and was 
measured on a scale which was photographed 
with the fish. For specimens in which the 
vertebrae were counted under the micro- 
scope, the length was measured directly with 
vernier callipers. 
RESULTS 
In a sample from Kaneohe Bay (10/25/47) 
in which the sexes were segregated, the mean 
vertebral number for 76 males was 42.618 and 
for 76 females, 42.645. The difference of 
0.027 had a standard error of 0.095. A dif- 
ference of this magnitude could arise readily 
from chance variation. It was concluded that 
sex need not be considered as a source of 
variation in mean vertebral number. This 
conclusion is in accord with those found by 
most workers in other species, e.g.. Tester 
(1937) in Pacific herring {Clupea pallasii) and 
Blackburn (1950) in the Australian anchovy 
{Engraulis australis), but it is in contrast with 
that found by Hart (1937) for the Pacific 
capelin {Mallotus villosus ) . 
In the Pacific herring (McHugh, 1942), 
there is a tendency for the mean vertebral 
number to be higher in larger fish, a pheno- 
menon which is presumed to be related to 
hereditary differences in the developmental 
rate of individuals. To investigate this possi- 
bility in the nehu, the fish of each sample were 
grouped in 3 -millimeter-length categories and 
the mean count for each was determined. The 
results, for length categories including 10 or 
more fish, are shown in Table 1 and Figure 
