sum of the individual epiphyseal fusion scores for 

 both flippers composed the FHpper Index (FI) for 

 that animal. When one of the flippers was dam- 

 aged, the score for the undamaged flipper was 

 doubled. 



The combined weights of the testes with the 

 epididymus removed were used as a measure of 

 sexual maturity in males. In this study, a pair of 

 testes was considered mature at 350 g. 



Just as menarche in human females does not 

 mean ovulation but the final developmental 

 stages for the ability to ovulate, delphinid ovarian 

 scars are here inferred to indicate ovulatory capac- 

 ity. The presence on the surface of at least one 

 corpus albicans or a corpus luteum indicated sex- 

 ual maturity. However, each ovary also was sliced 

 into sections 1 mm thick and examined for inter- 

 nal corpora. 



All statistical tests are described in Sokal and 

 Rohlf (1969). 



Results 



Development of right and left flippers did not 

 differ significantly in 27 males and 55 females 

 (P&0.05, /-test for paired comparisons). 



There was no significant difference (Ps0.05) in 

 weight between the left and right testes of 34 

 specimens (/-test for paired comparisons). 



In all but 2 of 25 mature cases, the left ovary had 

 more scars. 



Dentine layers are a poor indicator of sexual 

 development in D. delphis. The numbers of den- 

 tine layers and ovarian corpora are not sig- 

 nificantly related (P>0.10, Kendall's rank corre- 

 lation test). Both sexually mature and immature 

 females occur with 7-14 dentine layers (Figure 2). 

 Testes weights are so variable in the range of 7-12 

 dentine layers that they cannot be estimated (Fig- 

 ure 3), although significantly correlated over the 

 entire range of data (PsO.OOl, Kendall's rank cor- 

 relation). 



Body length is a poor indicator of sexual de- 

 velopment. Over body lengths 175-190 cm, testes 

 apparently undergo a transitional stage of growth. 

 Gonad weight cannot be accurately estimated 

 from body length over this range (Figure 4) al- 

 though the two are significantly correlated over 

 the entire range of data (PsO.OOl, Kendall's rank 

 correlation). Body length and ovarian scarring are 

 poorly correlated (P>0. 10, Kendall's rank correla- 

 tion). Body lengths 165-182 cm include both sex- 

 ually mature and immature females (Figure 5). 



DENTINE LAYERS 



Figure 2. — Ovarian corpora in relation to dentine layers in 

 Delphinus delphis. The stippled region indicates the range of 

 dentine layers over which sexually mature animals are indistin- 

 guishable from immature. 



1000 

 800 



T — I — I — I — I — I — I — I — I — I — I — I — I — I — r- 



I ' I ' I 



II * 

 I 



• • 



t • 



# 



DENTINE LAYERS 



FIGURE 3.— Testis weight in relation to dentine layers in Del- 

 phinus delphis. The stippled region indicates the range of den- 

 tine layers over which testes of mature and immature weights 

 overlap. 



The FI is significantly correlated with testes 

 weights (P=£0.001, Kendall's rank correlation) 

 although data are missing in a narrow range (Fig- 

 ure 6). However, inactive ovaries occur in a wide 



297 



