16 



KNOWLEDGE. 



[Janhary, 1903. 



dispute the palm of singularity with Pycnogotium, 

 since the latter has dispensed with a heart and second 

 appendages which the former retains. Both these genera 



Shiinchothoraj- mediterraneus, Costa. 9 From helow, legs omitted 

 From Dolirn. 



are included by Dolirn in the same family. Sars transfers 

 Rhynchothorax to his Pasithoidse, without commenting 

 on the diminished number of joints in its second aj)pen- 

 dages, or on the various eccentricities which seem to 

 demand for it a separate family, Rhynchothoracidse. 



There is still to be considered a genus which, if not the 

 most abnormal, is from some points of view the most 

 extraordinary of all the group. This is the genus of 

 the giant proboscis. It agrees with the Eurycydidae in 

 having ten-jointed second, and eleven-jointed third 

 appendages, but differs from them in the total absence 

 of chelifori, and iu not having its snout flexibly con- 

 nected with the trunk. The first species of it made known 

 to science was obtained within the Arctic Circle during 

 Parry's first voyage of discovery. The specimen was 

 " found at ebb tide on the shores of the North Georgian 

 Islands." This was described by Sabine in 1821 under 

 the name Phoxichihts proloscideus. It was subsequently 

 transferred to a new genus, Colossendeis, by Jarzynsky in 

 1870, and before this name had obtained currency it 

 successively received two others, Bhopalorhynchus* from 

 Wood-Mason, and Anomorliynchus from Miers, the former 

 implying that the proboscis is as massive as a club, and 

 the latter that it is what Dominie Sampson would have 

 called '-prodigious." The length of the adult Bhyncho- 

 thorax mediterraneus is a millimetre and a half, about a 



* This, liowever, is upheld as distinct bv G. H. Carpenter. 



sixteenth of an inch. Of Colossendeis proboscidea (Sabine) 

 the original description says : " proboscis more than twice 

 the lenglh of the body, being one inch eight-tenths, and 

 the body Ihree-cjuarters of an inch long." The length of 

 the leers is given bv Sabine as six inches, which would 



Colossendeis proboscidea (Sabine). Reduced. From Sars. 



therefore afford a total span of more than twelve. Sars 

 claims for the animal only a length of two inches with a 

 span of nine. But C. gigas, Hoek, from great depths in 

 the Southern Ocean, far surpasses either of these sets of 

 measurements, since the animal can attain a length of 

 over three inches with a span of more than twenty-four. 

 To display this creature life-size with its ambulatory limbs 

 unfolded and not overlapping, would require a double- 

 page plate of this magazine. The species of the genus are 

 numerous, and as they have generally been obtained from 

 considerable or very great depths, one may surmise that 

 there is some correlation between their length of limb and 

 the character of their abode. Where the sense of sight has 

 but scanty chances of employment, an extended organ of 

 touch may be very useful, and this will help to explain 

 the parallel case of deep-sea crustaceans, in some of which 

 the antennae seem ambitious of emulating telegraph wires. 

 Kevertheless, as it happens, it is the sliort-Unibed 

 C. hrevipes, Hoek, that is recorded from the deepest 

 station of all, 26.50 fathoms. The torrid temperature of 

 more than 70^ Centigrade, assigned to this abyss in Dr. 

 Hoek's " Challenger " report, is due to a slight confusion. 

 As may be seen from the "Narrative of the Cruise," 

 though the surface water at this station was moderately 

 warm, 70'8° Fahrenheit, the water at the bottom was only 

 a little above freezing point, 32'7^ F. Besides its colossal 

 blunt proboscis, Colossendeis has various other notable 

 characters. Meinert has remarked the retention of the 

 chelifori by the young of C. angiista, Sars, when these 

 juveniles were already more than half-an-inch long. Hoek 

 has drawn attention to the circumstance that the ovigerous 

 legs have not been found carrying ova in either sex. 

 These limbs attain a great length, and shew no sexual 

 difference. In the ambulatory limbs the terminal joint is 

 not here unguiform but awl-shaped, that is, more of a 

 needle than a hook. For what appear to be just reasons, 

 Pasithoe, Endeis, Rhynchothorax have been dismissed from 

 any close alliance with this genus. The family, therefore, 

 in which it stands alone, must take from it the title 

 Colossendeidae already accepted by various authors. 



That there is anything like finality in the classification 

 here adopted, with a few modifications, from Prof. Sars, is 

 not to be expected. All that can be safely assumed is that 

 this is at least as good as any of the several systems 

 hitherto offered, and iu the existing state of knowledge the 



