340 
PACIFIC SCIENCE, Vol. V, October, 1951 
differences between surveys are probably re- 
lated, in part at least, to differences in speed 
of tow. The average speed, as determined by 
readings of the Clarke-Bumpus current meter 
which was towed behind the boat in these two 
surveys, was 1,3 times as great in Survey 3 as 
in Survey 4. A high percentage of damaged 
eggs was also found in the two interim sur- 
veys when, in towing two small nets, the ship 
travelled faster than in towing a large and a 
small net. 
The percentage of agglutinated eggs, while 
ranging from 1.3 in Survey 3 to 11.0 in Survey 
2, did not seem to vary directly with speed of 
tow. Although the results are admittedly sub- 
ject to some degree of error in distinguishing 
damaged from agglutinated eggs, they sug- 
gest that about 5 per cent of the eggs were 
dead at the time of capture. The only other 
possibility is that they were damaged on 
capture at the start of the tow and that de- 
composition proceeded with sufficient ra- 
pidity to reduce them to the agglutinated 
condition in 10 minutes or less. This seems 
unlikely. 
If the agglutinated eggs are assumed to 
have been dead at the time of capture, the 
question arises as to the reason for their 
death. Some may have been regurgitated by 
chaetognaths and ctenophores, which were 
found to feed on the eggs. Others may have 
died “naturally” because of infertility, lack of 
vitality, or some other intangible factor. In 
this regard, D. Yamashita (unpublished) has 
observed that when normal undamaged eggs 
are kept in Petri dishes, a small percentage 
will die for no apparent reason. Determina- 
tion of the cause and extent of the egg mor- 
tality would be of considerable interest and 
importance in the study of the dynamics of 
the nehu population. 
LENGTH DISTRIBUTION OF LARVAE 
The length distribution of larvae taken by 
both large and small nets is given for each 
survey in Table 12. It will be observed that 
the length ranges from 2.5 to 12 mm. Only 
one individual of larger size '(20.5 mm.) was 
caught. The smallest nehu taken in the seines 
or night-light nets of the commercial fisher- 
men are about 23 mm. These are post-larval 
fish which have not yet completely meta- 
morphosed. It is clear, then, that the plankton 
nets did not sample the larva population 
completely. It seems reasonable to assume 
that the larger individuals escaped the net by 
virtue of their keener vision and greater 
swimming power, or these coupled with the 
adoption of the schooling habit. 
The average length of the larvae varies 
significantly from survey to survey; for Sur- 
veys 1 to 4, the average lengths are 4.99, 
5.73, 5.01, and 3.44 mm., respectively. The 
differences between surveys are probably re- 
lated to variation from day to day in the 
extent of spawning, presuming that several 
day groups are represented in the length 
distribution. Although the data for comparing 
sectors within surveys are scanty, there seems 
to be a general similarity in the length dis- 
tributions. 
The percentage frequency distribution was 
calculated for each survey and the resulting 
four percentages for each length group were 
averaged, giving the data of the last column 
of Table 12. The length frequency polygon, 
shown in Figure 6, is obviously multimodal, 
2. 3 + 5 6 7 e 9 10 II IZ 
STANDARD LENOTB - MM, 
Fig, 6. The per cent frequency distribution of 
larvae according to length, all data combined. 
