52 CRUSTACEA MALACOSTRACA. III. 



third segments distinctly longer than the others, and each considerably longer than the breadth of the 

 head; each process is shaped almost as in D. spinosum, terminating in two to four small teeth on the 

 truncate end (fig. 3 b), and a seta considerably shorter than in D. spinosum is inserted on or above 

 near the end ; each process has furthermore a long dorsal seta before the middle, and those of second 

 to fourth segments besides a seta beyond the middle and at least sometimes one or two proximal seta;. 

 All these processes and a distal part of the protuberances are adorned with a number of sharp gra- 

 nules (fig. 3 b). - ■ These segments have very long dorsal setae, each seta inserted on an elevated foot, 

 but each of the three anterior segments has only a single pair, placed near the posterior margin a 

 little from the middle (fig. 3 a), while the fourth segment has an anterior and a posterior pair, in the 

 large male even besides two pairs near the lateral margin. 



The three posterior segments have their lateral protuberances nearly longer than in I), spi- 

 nosum; the processes 011 fifth pair of coxae are slightly shorter than those on fourth segments (fig. 3 a), 

 but a little longer than those on sixth pair, and about three times as long as the processes on seventh 

 pair of coxae; all these processes are granulated like the anterior pairs. Each of these segments has 

 a single pair of dorsal long hairs. 



The thoracic legs, as far as is known (fig. 3 a), seem to be somewhat longer than in D. spino- 

 sum; the prehensile hand (fig. 3 d) is a little more slender than in that species. 



The abdomen shaped as in D. spinosum, generally with 3, in the large male with 4 pairs of 

 marginal spines. The end of abdomen of one specimen has five minute teeth (fig. 3 e), but in other 

 specimens the end has only a single tooth or two or three obsolete, irregular teeth. I have been 

 unable to point out any difference in the shape of the opercula in either sex between this species and 

 D. spinosum (uropods lost). 



Length of the large male 2'6 mm., while the other specimens are smaller, most of them 2 — 2-5 mm. 



Remarks. D. paradoxum is easily separated from D. spinosum by the thoracic processes, 

 which are conspicuously set with a good number of granules quite wanting in the other species. 

 Furthermore the processes of the four anterior segments are longer than in D. spinosum, the processes 

 on the coxa; of fifth segment almost as long as the preceding pair of processes, while in D. spinosum 

 they are quite short; in D. paradoxum the coxal processes decrease much in length from fifth to 

 seventh pair, while in D. spinosum they increase in length backwards. The existence of a pair of 

 setiferous tubercles on the three posterior segments is also a valuable character, and the relative length 

 of the joints in the antennular and antennal peduncles affords other characters, when these appendages 

 have been preserved. 



Occurrence. Taken by the "Ingolf" at a single station. 



South-West of Iceland: Stat. 78: Lat. 6o°37' N., Long. 27°52' W., 799 fath., temp. 4-5; 20 spec. 



It may be noted that enormous quantities of sponges were procured at that Stat. 78. Among 

 the animals found were the specimens of I>. paradoxum, a specimen of D. spinosum, furthermore the 

 two specimens of the above-described Pseudomunna hystrix, specimens of the extremely long-legged 

 Pseudotanais longipcs H. J. H., and of Typhlotonais eximius H. J. H., which possesses very elongate and 

 slender chelae. I suppose that all these animals live on sponges, and that there may be some correla- 

 tion between their armature or very elongate appendages and their mode of living. 



