336 ALLAN HANCOCK PACIFIC EXPEDITIONS VOL. 2 



are few eggs none of which are posterior to the genital pore. Contrasted 

 with this, Nagaty's material is from marine fishes, the mouth is almost 

 always well posterior to midbody, the testes are tandem; the ovary al- 

 ways anterior to the anterior testis; the cirrus sac is longer than % 

 body length; the utems rarely extends anterior to the vitellaria and is 

 better developed posterior to the mouth, while eggs can occur posterior 

 to the genital pore in the Red Sea specimens (see Nagaty's fig. 8) and 

 usually do so in my American specimens. The most constant difference 

 is the anterior position of the ovary which is never lateral to the anterior 

 testis. The next most important difference is probably the length of the 

 cirrus sac. Both the anterior and posterior extent of the uterus is less 

 in j^oung specimens. In the writer's experience gasterostomes either are 

 specific or tend to show host specificity. This tendency indicates that the 

 Bucephalus from the related marine hosts is likely to be distinct from 

 B. polymorphus. The marine species is therefore renamed B. varicus, a 

 name indicating its variability. 



Nagaty, impressed by the variability of the tentacles, considered B. 

 elegans Woodhead, 1930 a synonym of B. polymorphus. The writer does 

 not agree with this conclusion. The eggs of B. elegans are approximately 

 twice the size of those of B. polymorphus. 



The American specimens of B. various do show considerable varia- 

 tion, some of which is probably due to degree of contraction, some to 

 degree of maturity, and some perhaps to host influence. Egg size was 

 rather constant within a specimen but varied considerably in different 

 specimens, especially from different hosts. But eggs from the Pacific host 

 were intermediate between extremes found in different Atlantic hosts. 

 Thus, Pacific specimens have eggs somewhat wider than those from 

 Caranx ruber at Tortugas, but the difference is not deemed significant 

 considering all variations. All the variations of the tentacles were noted 

 as by Nagaty with the additional occasional occurrence of a third small 

 ventral process near the tip of the tentacle (fig. 6). The writer interprets 

 the threadlike process on the tips of some tentacles as a product of dis- 

 integration and believes the slightly elevated bosses to represent com- 

 pletely retracted tentacles. 



The following diagnosis is based on specimens from both the Atlantic 

 and the Pacific but does not include Nagaty's data which, however, agree 

 in all essential points. Although the collections from the three hosts are 

 considered to be one and the same species, some of the measurements are 

 separated for purposes of comparison. 



