334 
PACIFIC SCIENCE, Vol. V, October, 1951 
TABLE 10 
Number of Eggs and Larvae according to Haul Order, and Depth for Survey lb in the Vicinity of 
Station 4 
HAUL 
DEPTH OF TOW 
NUMBER OF EGGS 
NUMBER OF LARVAE 
SURFACE 
DEEP 
SURFACE 
DEEP 
SURFACE 
DEEP 
feet 
feet 
I A 
2 
19 
79 
61 
1 
1 
IB 
1 
16 
98 
93 
1 
1 
II A 
2 
19* 
19 
5* 
- 
- 
IIB . . . 
1 
16 
51 
49 
3 
4 
IIIA 
2 
20 
3 
2 
1 
2 
IIIB 
1 
17 
19 
26 
4 
4 
IVA 
1 
17 
191 
133 
- 
- 
IVB 
1 
16 
146 
124 
- 
1 
VA 
1 
17 
120 
84 
- 
1 
VB 
2 
18 
141 
97 
- 
1 
VIA 
2 
17 
58 
52 
6 
1 
VIB 
2 
18 
53 
51 
2 
1 
Totals 
978 
111 
18 
17 
* Net hit bottom; some damage to net and loss of plankton. 
the results may be most plausibly explained j 
on the basis of non-random horizontal dis- 
tribution of the eggs. 
To explain the foregoing results satis- 
factorily it is necessary to assume only that 
there is a focus of abundance of eggs in the 
vicinity of Station 4, with gradients of de- 
creasing abundance extending outward in all | 
directions. The nature of the egg distribution 
which is postulated to have occurred during 
Surveys la and 1^ is illustrated in Figures 2 
and 3. Transection of the area at various dis- 
tances from the focus would cause the great 
variability in catch and would readily account 
for the observed significant differences and 
interactions. As will be shown later, the pres- 
ence of a focus of abundance of eggs in the 
vicinity of Station 4 is also demonstrated by 
the results of the general surveys. 
It seems likely that non-random distribu- 
tion of the eggs occurs at all stations. Slightly | 
different gradients of abundance are passed ! 
through in making A as compared with B 
hauls, often causing large differences in the j 
number of eggs caught. A similar situation ! 
may also occur in the case of larvae, although j 
the differences between A and B hauls are, in 
general, less pronounced and persistent. Be- | 
As shown by the analysis of the data 
(Table 11), there are significant differences 
between substations, which are due mostly 
to the large difference between series. The 
difference between the (geometric) means for 
haul order (A, 35; B, 63) is not significant: in 
Series 1 the means differed to some extent 
(A, 13; B, 42), whereas in Series 2 they were 
similar (A, 96; B, 94). However, there is a 
significant interaction, now designated as' 
SXH (substations by hauls), showing lack of 
consistency in the differences between A and 
B hauls from substation to substation. Again, 
TABLE 11 
Analysis of Variance of Egg Catches in Survey 
lb ACCORDING TO SERIES, SUBSTATIONS, HAUL ORDER, 
AND Depth 
SOURCE OF 
VARIATION 
DEGREES OF 
FREEDOM 
SUM OF 
SQUARES 
MEAN 
SQUARE 
Substations (S) 
5 
4.91036 
0.98207* 
Between series 
1 
2.60827 
2.60827 
Within series 
4 
2.30209 
0.57552 
Haul order (H) 
1 
0.38457 
0.38457 
Depth (D) 
1 
0.07549 
0.07549* 
SXH 
5 
0.81374 
0.16275** 
SXD 
5 
0.04786 
0.00957 
HXD 
1 
0.04437 
0.04437 
SXHXD 
5 
0.05212 
0.01042 
* Significant. 
** Highly significant. 
