328 
PACIFIC SCIENCE, Vol. V, October, 1951 
TABLE 4 
Number of Nehu Eggs per 10-Minute Surface Haul with the 22 cm. Net (Surveys 1, 3, and 4) or the 
12.3 CM. Net (Survey 2), according to Survey, Station, and Direction (A and B), and Summary 
STATION 
SURV 
'EY 1 
SURV 
EY 2 
SURV 
EY 3 
SURV 
EY 4 
MEAN PER CENT 
OF SURVEY 
TOTAL 
A 
B 
A 
B 
A 
B 
A 
B 
1 
— 
— 
— 
— 
— 
— 
1 
— 
0.17 
2 
- 
- 
- 
- 
- 
1 
- 
- 
0.04 
3 
37 
83 
3 
6 
6 
22 
- 
2 
7.02 
4 
244 
193 
7 
26 
217 
279 
11 
15 
46.14 
5 
130 
15 
27 
20 
5 
- 
13 
17 
21.52 
6 
2 
1 
- 
- 
- 
- 
- 
- 
0.07 
7 
- 
- 
- 
- 
- 
- 
1 
- 
0.17 
8 
1 
- 
- 
- 
- 
- 
4 
3 
1.19 
9 
6 
7 
1 
— 
— 
14 
10 
1 
3.05 
10 
— 
— 
1 
1 
— 
— 
— 
— 
0.54 
11 
- 
- 
- 
- 
- 
- 
- 
- 
0.00 
12 
1 
5 
- 
- 
1 
- 
- 
- 
0.20 
13 
4 
28 
- 
- 
- 
10 
- 
- 
1.25 
14 
26 
9 
- 
- 
- 
- 
20 
6 
5.22 
15 
12 
19 
- 
- 
1 
1 
5 
6 
2.70 
16 
1 
- 
- 
- 
1 
3 
9 
7 
2.87 
17 
72 
18 
— 
1 
3 
2 
6 
4 
4.43 
18 
29 
8 
— 
— 
— 
— 
— 
— 
0.93 
19 
1 
- 
- 
- 
- 
- 
- 
1 
0.19 
20 
22 
16 
- 
- 
- 
- 
2 
5 
2.13 
21 
- 
- 
- 
- 
- 
- 
1 
- 
0.17 
22 
- 
- 
- 
- 
- 
- 
- 
- 
0.00 
23 
— 
— 
— 
— 
— 
— 
— 
— 
0.00 
Totals ... 
588 
402 
39 
54 
234 
332 
83 
67 
100.00 
of the myotomes, the position of the anus, 
the projecting lower jaw (up to about 15 
mm.), and the lateroventral position of the 
pectoral fins. 
Many of the larvae taken in the tows were 
damaged to a varying extent. In young larvae 
the yolk was often torn from the yolk sac, 
and the fin fold was usually shredded. In 
older larvae part of the head, including the 
eyes, was often missing. This complicated the 
problem of identification. More serious, how- 
ever, is the possibility that some nehu larvae, 
broken into fragments, were not recognized 
as such and were not included in the counts. 
VARIABILITY OF THE DATA AND RELATIVE 
EFFICIENCY OF THE NETS 
It will be recalled that, in sampling, two 
nets were hauled simultaneously, first in A 
direction and then in B direction at each sta- 
tion. Following the method of Winsor and 
Clarke (1940), the data may be used to in- 
vestigate the components of variance and the 
relative efficiency of the nets. Analyses were 
made for the egg catches of each survey, using 
three criteria of classification— stations, haul 
order, and nets. To approximate normal dis- 
tributions, the data were transformed to 
logarithms. The method of analysis is ex- 
plained by Snedecor (1948: 11.14). Unfor- 
tunately, because of the small numbers taken, 
it was necessary to omit data from many of 
the stations. The stations included are as 
follows: Survey 1 — 3, 4, 5, 13, 14, 15, 17, 18, 
and 20; Survey 2—3, 4, and 5; Survey 3—3, 
14, 16, and 17; and Survey 4—4, 5, 8, 9, 14, 
15, 16, 17, and 20. The results are given in 
Table 7. 
