764 



Fishery Bulletin 91(4). 1993 



CH NZ PE SA 



Orbit length 



CH NZ PE SA 



Antorbitel procitt length 



CH NZ PE SA 



Internal mm width 



CH NZ PE SA 

 Pterygoid length 



CH NZ PE SA 

 Upper tooth row length 



CH NZ PE SA 



Lower tooth row length 



CH NZ PE SA 

 Remus length 



CH NZ PE SA 

 Remus height 



CH NZ PE SA 

 Tooth count upper lert 



CH NZ PE SA 



Tooth count upper right 



CH NZ PE SA 

 Tooth count low«r l«ft 



Figure 2 (Continued) 



CH NZ PE 3A 

 Tooth count uppar left 



P<0.0001), presence of an accessory foramen above the 

 foramen magnum (variable 16, P=0.02), and the de- 

 gree of contact, posteriorly, between pterygoid hamuli 

 (variable 23, P=0.03). 



Upon comparison of frequencies for variables 16 and 

 23, two similarity pairs became apparent: New 

 Zealand/southwestern Africa and Peru/Chile. Within 



each subgroup the monomorphic status for both char- 

 acters was confirmed (chi-square, df=l, P>0.41 in four 

 tests). Southwest African L. obscurus differed very sig- 

 nificantly in character 1R from all other stocks. By 

 removing the African subsample, frequencies for 1R in 

 the three remaining geographic units became homoge- 

 neous (chi-square=0.58; df=4, P=0.96). 



