FISHERY BULLETIN: VOL. 71, NO. 2 

 Table \. — Characteristics distinguishing Hepomadiis glacialis (holotype) from H. tener. 



Feature 



H. glacialis (holotype) 



H. lener 



Dorsum of carapace 



Postrostral carina 



Setae on endite of 

 basis of second 

 maxilla 

 Upper side 



Lower side 



Strongly arched to postcervicol 

 sulcus 



Interrupted, disappearing on cardiac 

 region, becoming distinct again and 

 extending almost to posterior margin 

 of carapace 



5 and 6 strong, 10 and 4 weak (right 

 and left endite, respectively) 



5 and 9 strong, 3 and 5 weak 



Almost straight or very slightly 

 arched, convexity extending beyond 

 postcervicol sulcus 



Uninterrupted, extending posteriorly 

 about 0.8 length of carapace 



0; 2-4 strong, 1 -4 weak 



0; 2-5 strong, 1-4 weak 



value in differentiating two species of western 

 Atlantic Hepomadus, I have divided the ma- 

 terial into three somewhat arbitrary groups 

 based on the shape of the midposterior margin 

 of the fourth and fifth pleonic terga: Group 1, 

 straight; Group 2, slightly produced into a 

 short, sometimes barely perceptible, projection; 

 and Group 3, produced into a conspicuous 

 spine. Intermediate stages (Group 2) between 

 the two extremes (Groups 1 and 3) have not 

 been discussed previously, although Bouvier 

 (1908), in his description of a male H. tener 

 from the Sargasso Sea, indicated that the 

 dorsal carina of the fourth and fifth pleonic 

 terga is produced into a small projection. 



As stated above, Burkenroad indicated that 

 in the specimen reported by Milne Edwards 

 and Bouvier, which fits into Group 3, the 

 interval between the last two rostral teeth is 

 smaller than in the specimens with straight 

 margins (Group 1) he had studied. I have 

 found that this tooth interval varies consider- 

 ably, from 0.30 to almost 0.60 of the distance 

 between the first and the second tooth, and 

 that the variations are not correlated with the 

 shape of the post^^rior margin of the pleonic 

 terga (Figure 9). It is true that according to 

 the figure of Milne Edwards and Bouvier, in 

 their Group 3 male (17 mm c.l,), the interval 

 is 0.30, whereas in the Group 2 male described 

 by Bouvier, the interval is 0.55 [according to 

 his Figure 2, on plate 13], and in the female 

 with 47 mm c.l., from off Virginia, belonging 

 to Group 1, which Burkenroad had studied, 



the interval is 0.50. However, in the male 

 holotype of H. tener (17.5 mm c.l., Group 1), 

 which Burkenroad had not seen, the interval is 

 0.30, as in the 17 mm c.l. male of Group 3. 

 Furthermore, among the material I have ex- 

 amined, two females of Group 3 of about same 

 size (36 mm c.l. and 37 mm c.l.) possess a tooth 

 interval of 0.40 and 0.60. Finally, in a small 

 female of 14.5 mm c.l. and in another of 47 

 mm c.l., both belonging to Group 2, the tooth 

 interval is about the same, 0.35 and 0.40, 

 respectively. The data presented herein demon- 

 strate that there are no significant differences 

 in the tooth interval between the three groups. 



The available information is too meager — 

 particularly on animals of Group 1 — to deter- 

 mine whether or not a clear correlation exists 

 between the relative length of the rostrum and 

 the shape of the midposterior margin of the 

 fourth and fifth pleonic terga. Nevertheless, 

 the ratio of the length of the rostrum to the 

 length of the carapace in the specimens examined 

 by me are presented in Table 2. 



The relative length of the posteromedian 

 spine on the third pleonic tergum varies greatly, 

 from 0.3 of, to equivalent to, the distance 

 between the transverse groove and the posterior 

 margin of the tergum (adjacent to base of 

 spine). The correlation of this ratio with in- 

 crease in size of the animals seems obscure, 

 and the correlation of the relative length of the 

 spine with the shape of the midposterior margin 

 of the fourth and fifth pleonic terga does not 

 appear to indicate the presence of more than 



450 



