ANDERSON, GEHRINGER, and BERRY: NUMBERS OF VERTEBRAE AND LATERAL-LINE SCALES 



including the triangular-shaped last or ultimate 

 centrum articulating with the hypural bones. 

 All counts were made from X rays. Vertebral 

 counts for individual specimens are used twice 

 in correlations with paired (bilateral) scale counts. 

 Specimens with obvious abnormalities in verte- 

 bral structure were not included in the tabu- 

 lations. 



Pored lateral-line scales — counted from the 

 first pored scale in the lateral line to and in- 

 cluding the last pored scale in the lateral line, 

 lateral to the median base of the caudal-fin rays. 

 Specimens with one or more scales missing from 

 the lateral line were included in the tabula- 

 tions, if the position of the scale pockets or the 

 remaining scales were adequate to allow an 

 accurate reconstruction. Bilateral counts were 

 made for each specimen, and the counts for 

 both sides of each fish were included in all 

 tables and in the correlations with vertebrae. 

 The term scale or scales refers specifically to 

 pored scales in the lateral line. 



RESULTS 

 Vertebrae 



Vertebrae in our samples of the 10 western 

 Atlantic species range from 44 to 62. We know 

 of one higher count for a synodontid — the holo- 

 type of Synodus ulae from Hawaii has 64 verte- 

 brae. Statistics based on vertebral samples for 

 the western Atlantic species are listed in Table 1 

 and graphically depicted in Figure 1. Samples 

 of at least six of the species are probably too 

 small (and possibly too heterogeneous) to 

 adequately define the ranges of vertebral comple- 

 ments for these species. This is obvious from com- 

 parison of skewedness of the range and/or stan- 



dard deviations and the relatively large values 

 of the standard errors. For the two species for 

 which we have the largest and most reliable 

 samples, Sy. foetens and Sy. intermedius, the 

 relation of variance to meristic complement is 

 similar to that we have noted in some other fish 

 families — the species with the larger number or 

 larger range of elements has the greater variance. 

 The extremely high variance and extensive 

 range of vertebrae in Saurida caribbaea suggest 

 a taxonomic or ontogenetic problem as yet 

 unidentified. 



Intraspecific geographic variation in these 

 lizardfishes has not been investigated, but our 

 limited samples of four of the species suggest 

 that it may exist. Synodus intermedius appears 

 to have fewer vertebrae in tropical continental 

 waters and more in insular areas than in tem- 

 perate and subtropical continental waters. 



Area 



n 



These samples are relatively homogeneous, and 

 the means of the three samples are significantly 

 different (P < 0.01, anova). Synodus foetens 

 may have a parabolic cline in mean vertebral 

 number, with lower average numbers in tropical 

 areas and higher average numbers to the more 

 temperate north and south. 



Area 



n 



Table 1. — Statistics from numbers of vertebrae, numbers of pored lateral-line scales, and correlation coefficients 

 of the two variates from samples of the 10 sptecies of western Atlantic Synodontidae. 



203 



