286 BULLETIN OF THE BUREAU OF FISHERIES. 



recently by Naef (19126, p. 199)- Wiilker has likewise reported under the same name a species of the 

 genus from Japan (1910, p. 5). Not having any material from other regions available for comparison, 

 I have felt constrained to follow the precedent thus made for me, especially since I have not been able 

 to seize upon any characters of sufficient definiteness to separate the Pacific species from the Atlantic. 

 Nevertheless it should be remembered that our knowledge of the group is still woefully incomplete, 

 and a careful account is needed of all the diverse stages which specimens from the various regions pass 

 through during their life history, before absolute confidence can be felt in such wholesale relegation to 

 the synonymy. 



Should it eventually appear that there are more species of Tremoclopus than recent writers are 

 inclined to admit, I have no doubt whatever that the Hawaiian specimens are identical with the form 

 described by Souleyet as Octopus gracilis. Tryon (1879, p. 131) gives the following brief paraphrase of 

 his very inadequate description: "Body rounded, smooth; head small; eyes large, prominent; arms 

 graceful, the upper very long, lower short, connected by a web. Phosphorescent and with metallic 

 reflections when living. Length 24 mm. " 



Certainly these few lines contain little which can serve either to prove or to disprove the identity 

 of Souleyet 's specimen with those obtained by the Albatross, but upon turning to the figure we 

 find a representation which accords very well with Hawaiian specimens of the same size, and as both 

 are from the same general oceanic region and the habit of the animals is admittedly pelagic, we find at 

 once a strong a priori reason for uniting them. From the present specimens the figure of 0. gracilis 

 differs most strikingly in its more inflated body (a character of trifling importance) and somewhat wider 

 web between the ventral arms. The eyes are also represented as being semipedunculate, but I suspect 

 that this appearance may be due to the loss of the delicate integument which usually envelops the head. 

 The proportions of the arms are, however, much the same, and it is further noteworthy that a specimen 

 presumably of this species taken by the Challenger in the western Pacific differs in no described char- 

 acter from the Albatross specimens (Hoyle 1886, p. 71). 



With the same form I would be also disposed to unite the Octopus dubius, also of Souleyet, 

 especially since the most immature of the specimens before me is almost a counterpart of his figure 

 and is of similar dimensions (6 mm.). 



The two female specimens dredged by the A Ibatross in the tropical Pacific, which Hoyle (1904, p. 12) 

 referred with some doubt to T. quoyanus, likewise agree perfectly in all the characters stated, have the 

 same curious arrangement of the chromatophores, and are surely conspecific. The single poorly pre- 

 served individual taken by the Challenger in the south Pacific (Hoyle 18S6, p. 70) also is probably 

 identical, so that on the whole it can be affirmed with certainty that few Hawaiian cephalopods exhibit 

 a wider distribution. 



The species is one of the most beautiful and graceful of all cephalopods. 



Family ALLOPOSID.£ Verrill 1881. 



AUoposidm Verrill 1881, p. 365. 



ATyuiiautid<E (pars) Naef r9rzb, p. 197, 198. 



Sexes dimorphic, the female larger than the male and often attaining considerable dimensions. 

 Body soft and gelatinous, aquiferous pores lacking. Arms united throughout the greater portion of 

 their length by an extensive umbrella; suckers in 1-2 rows. Hectocotyiization affects the whole of 

 the right third arm, which has its origin in a specialized sac in front of the eye, and when mature is most 

 probably caducous. 



Genus ALLOPOSUS Verrill 1880. 



Haliphron Steenstrup 1859, p. 183 (nomen nudum, fide Hoyle). 

 Alloposus Verrill 1S80, p. 393. 

 Alloposus Verrill 1SS1, p. 365. 

 Alloposus Hoyle 1886. p. 72. 

 AIloposus Naef 1912b, p. 198. 



This being the only known genus, its characters are entirely those of the family. 

 Type. — Alloposus mollis Verrill 1880 (monotypic), originally described from off the New England 

 coast. 



