56 



FISHERY BULLETIN OF THE FISH AND "WILDLIFE SERVICE 



crease. To determine whether the latter was true, 

 comparisons were made of the ova-diameter fre- 

 quency polygons for 23 fish in stage F according 

 to the months in which they were collected. The 

 number of ova larger than 0.55 mm. in diameter 

 was compared with the number of those measur- 

 ing between 0.18 and 0.53 mm. Ratios of the 

 numbers of ova comprising intermediate and ma- 

 turing groups for all fish in stage F are shown in 

 table 3. It may be seen that in June the ratios 



ranged between 4.36 : 1 and 5.98 : 1, the average 

 being 5.08 : 1. For July the average ratio, 4.50 : 1, 

 was slightly less. It declined more in August 

 (3.80: 1) and still more in September (3.25: 1). 



To determine the consistency in each month of 

 the ratios obtained for fish with ovaries in stage 

 F, the chi-square test was used. No significant in- 

 teraction (the smallest probability value obtained 

 was about 0.70) was found among the individual 

 samples collected during any one month; there- 



Table 3. — Ratios of ova between 0.18 and 0.53 mm. to 

 ova larger than 0.55 mm. for all fish with ovaries in 

 stage F, Jane-September 1950 



fore, the pooled data for each month were used 

 to test the consistency of the ratios among the 4 

 months, June through September. As shown by 

 an interaction chi square of 27.72 (P<0.01), ac- 

 cording to the method of Snedecor (1946, sec. 

 9.10), from the data presented in table 4, it is 

 unlikely that the monthly ratios were drawn from 

 a homogeneous population. Thus, there is a sig- 

 nificant decline in the ratios of the maturing to 

 the intermediate group through successive months. 

 This is not due to a change in the size of fish 

 during the season. It is seen in table 3 that the 

 mean weights of yellowfm tuna with ovaries in 

 stage F did not change appreciably during the 

 period considered. By an analysis of variance, it 

 was found that among these samples, the difference 

 between months was not more significant than the 

 difference within months (2?= 1.31; P>0.05). 

 In other words, the population did not change 

 significantly as to size during the season. Thus, 

 if the change is not connected with size of fish, it 

 doubtless reflects a decline in proportionate num- 

 bers of ova composing the intermediate groups. 



