FISHERY BULLETIN: VOL. 72. NO. 1 



housed in the BPBM (3398). Measurements of 

 the 106.8 mm SL male holotype (Fowler gave 

 an undefined length of 134 mm) expressed as 

 percent of standard length are given in Table 

 1. We obtained the following counts on this 

 specimen: dorsal rays 86; anal rays 62; caudal 

 rays 16; pectoral rays (both sides) 12; lateral- 

 line scale rows 130; pored lateral-line scales 

 83; gill rakers 19 (5 + 14). Prowler (1923) gave 

 counts of 132 scale rows and 20 gill rakers (6 

 -I- 14). Otherwise, our counts agree with his. 



Photographs of recently collected female 

 (106.4 mm SL) and male (114.4 mm SL) speci- 

 mens (both BPBM 14102) are shown in Figures 

 1 and 2. 



Counts 



Dorsal and anal ray counts, expressed as 

 bivariate relations, for 101 specimens are given 

 in Table 2. Dorsal rays ranged from 84 to 95 

 {X — 87.9) and anal rays ranged from 64 to 

 70 (A' = 66.3). Although the dorsal ray counts 

 are skewed to the right, the distribution does not 

 deviate significantly from a normal distribution 

 (P>0.2; Kolomogorov-Smirnov test for good- 

 ness of fit, D = 0.0702). 



There were considerably fewer pored lateral- 

 line scales (Table 3) {X = 80.9) than the numbe£ 

 of vertical scale rows above the lateral line (X 

 = 131.6). 



There was a tendency towards more pectoral 

 rays (Table 3) on the ocular side (X = 12.24) 

 than on the blind side {X = 11.49). The upper 

 pectoral ray on the ocular side is reduced and 

 sometimes inconspicuous. The upper pectoral 



ray on the blind side is also reduced, but easily 

 visible. 



BotJiKs thompso)ii appears to be unique among 

 si)ecies of BotJius in usually jjossessing 16 caudal 

 rays. Of 163 specimens examined (Table 3), 2 

 (1.2% ) had 15 rays, 157 (96.3% ) had 16 rays, and 

 4 (2.5% ) had 17 rays. The caudal rays are usually 

 associated with the four hypural elements. Of 

 31 cleared and stained specimens having 16 

 rays, 11 had a caudal ray formula of 4-4-4-4 

 (dorsal element counted first). Other formulae 

 obtained and number of specimens are as follows 

 (rays articulating between elements are enclosed 

 by parentheses): 4-4-3-(l)-4, (6 specimens); 

 4-4-3-5, (5 specimens) ; 3-( l)-4-4-4, (4 specimens) ; 

 3-(l)-4-3-(l)-4, (2 specimens); 3-(l)-4-3-5, (1 

 specimen); 3-(l)-4-3-5, (1 specimen); 4-4-3-5, 

 (1 specimen). Two specimens with 15 rays 

 had formulae of 4-4-3-4 and 3-(l)-4-4-3. 



None of the caudal rays of B. thompsoiii was 

 associated with the neural and haemal spines 

 of the penultimate vertebra or articulated in the 

 space between the spines and hypural elements. 

 Thus, B. thompsoni differs from certain other 

 species of Botlius which have rays associated 

 with the neural and haemal spines of the penul- 

 timate vertebra or which occur in the interspace 

 between the spines and hypural elements. Gutherz 

 (1970) gives a formula of 1-4-4-3-4-1 for larval 

 BotliHs (species not determined) from the 

 western North Atlantic. We obtained the same 

 formula for 12 cleared and stained specimens 

 of B. pa)ithe7HHUs (Riippell) from the Hawaiian 

 Islands. In these specimens the first and last 

 rays most often articulated in the interspace 

 between the spines and hypural elements. These 



Table 1. — Bothiis thompsoiu: Measurements of 13 characters for holotype, 29 males, and 31 females expressed as percent 

 of standard length. Holotype excluded from regression statistics (a = ordinate intercept, h = regression coefficient). 



238 



