so NATURAL SCIENCE. July, 



containing otoliths, are often abundantly developed, but there are no 

 eyes. The preceding characters appear to be those in which the 

 Elasipoda have been modified for an abyssal existence, and they 

 show that this family has had no connection with the others, at all 

 events with the Synaptidae, Molpadidae or Dendrochirotae, since a 

 remote period. 



The following characters appear to have been transmitted without 

 change from their primitive ancestors. The calcareous deposits, both 

 in the perisome and in the spicular ring surrounding the gullet, singu- 

 larly resemble those of larval forms. The water-vascular system, as 

 in the larvae only of other holothurians, but as in the adults of other 

 classes of Echinoderma, is often in persistent communication with the 

 exterior, and that, too, not only by one pore, but sometimes by a 

 number of pores closely crowded together so as to form a kind of 

 external madreporic tubercle. 



Despite the archaic and distinctive characters of the Elasipoda, 

 one of their families, the Psychropotidae, closely resembles some of 

 the Aspidochirotae, such as Stichopus, and especially Pcelopatides — a 

 genus first found by the " Challenger " ; and this suggests that 

 Aspidochirotae and Elasipoda sprang from a common branch. Thus 

 confirmation is lent to the view that the ancestors of the Holothurians 

 were pedate, with an open stone-canal and a well-developed ambu- 

 lacral system. 



Considering the habitat of the Elasipoda, it can hardly be 

 supposed that they undergo development with metamorphosis through 

 a free larval stage. That direct development is possible, was shown 

 by the interesting observations made during the "Challenger" 

 Expedition on the development of some shallow-water Holothurians, 

 viz., Cladodactyla crocea, from Stanley Harbour, and Psoitis ephippifcr, 

 from Heard Island. In the females of the former the young were 

 closely packed and adhering to the dorsal pedicels, while in the latter 

 PI. X., Fig. 4) the embryos were developed in a kind of marsupium 

 formed by the plates of the upper surface. There is little doubt that 

 the eggs are impregnated in the ovary, and that the free larval stage 

 is omitted. 



HjALMAR ThEEL. 



Annelida and Nemertea. 



The collection of Annelida made by the " Challenger " was by- 

 far the largest hitherto brought together as regards both individuals 

 and species, no less than 330 different forms having been procured, 

 and the majority, viz., about 220, were new to science. In com- 

 parison, the collection made by the American ship " Blake" contained 

 only 102, and the extensive Pliilippine series of Semper about 160. 

 Though no new family had to be constituted, yet in every one new 

 species, and in many new genera, occur — some of them of a remark- 

 able character. Further, this fine collection showed that the general 



