56 



Fishery Bulletin 90(1). 1992 



that 6 of the original 36 measurements (i.e., W. Lf. 

 Premax. [at midline of Nares], W. Rt. Premax. [at 

 midline of Nares], Separation of Pterygoids, L. Lf. 

 Tympanic Cavity, L. Rt. Tympanic Cavity, and W. at 

 Pterygobasioocipital Sutures; abbreviations used in 

 these and other character names are listed in footnote 

 a of Table 1) should be deleted, because we were not 

 able consistently to repeat these measurements. For 

 some other measurements, there were differences 

 between investigators, but the differences were con- 

 sistent (e.g., one obtained measurements that were 

 smaller than those reported by the other). Therefore, 

 we calculated regression equations for each of the 

 remaining characters based on the 81 jointly-measured 



specimens. These regression equations were used to 

 convert the measurements from the rest of the initial 

 specimens to appropriate values for inclusion with the 

 measurements taken by Perrin. Through these pro- 

 cedures, we developed a data set of 30 characters (listed 

 in Table 1) for 246 specimens. 



Only specimens that were largely complete were in- 

 cluded in the analysis. Missing values (1.34% of total) 

 were estimated by linear regression ("Missing Data 

 Estimator" program developed by Dennis M. Power, 

 Santa Barbara Mus. Nat. Hist., pers. commun.) onto 

 the character that explained the greatest proportion 

 of the variance for the variable under consideration. 



Specimens then were assigned to 5° latitude-longi- 



