REPORT ON THE HOLOTHURIOIDEA. 145 



are by far most commonly seen in the deep-water Holothurioidea in question. 

 Supposing that the four arms of these deposits, instead of being free and independent, were 

 connected at their ends with one another, each spicule would give origin to a plate with 

 four holes, representing the first stage in the development of a plate or wheel. The 

 process by which the plates and wheels of Elasipoda are developed from a spicule has 

 been already sufficiently explained in the foregoing anatomical description. Besides, I 

 may be justified in comparing the wheels of the Elasipoda with those of the larva of 

 Synapta, but, as they are constructed in a different manner, the resemblance which the}' 

 present is more apparent than real. In fact, all the wheels of the Elasipoda have the 

 nave perforated by a large hole, from the edge of which rises a crown of four to six arcuate 

 rods, while, on the contrary, the wheels of the larva of Synapta, in conformity with those 

 of other Apoda, as, for instance, Chirodota, Trochoderma, and Myriotrochus, are devoid 

 of a central hole as well as of a crown. The only exception to this rule occurs in the 

 very strange minute hat-shaped bodies in Elpidia glacialis, which, however, by possess- 

 ing a central crown composed of three rods, seem to approach more nearly to the wheels 

 of the Elasipoda 1 than to those of the Apoda. 



According to Miiller, Baur, 2 Metschnikoff, 3 &c, the first traces of a calcareous ring in 

 the larvae appear as separate unbranched spicules surrounding the oesophagus. As the 

 spicules grow larger, their ends become bipartite and gradually dichotomous ; finally, the 

 spicules become connected with one another so as to form a complete ring. On compar- 

 ing the larval ring with that of the Elasipoda, some very singular similarities present 

 themselves. In fact, the whole family Elpidiidae is distinguished by possessing a cal- 

 careous ring composed of spicules, which strikingly remind one of those of the larvae, 

 excepting that the branches or arms, which radiate from their ends, are more outgrown. 

 However, it is of importance to remember that the ring is made up of only five radial 

 pieces, while the larvae, as it seems, have commonly ten, five radial and five interradial, 

 the former five being probably first developed. The five spicules of the Elasipoda being, 

 with a few exceptions, separated from one another, the resemblance becomes more striking. 

 In the Deimatidae the ring is in a somewhat more advanced state of development, the 

 spicules having been converted into a fragde spongy network. 



The larvae of the Apoda and Pedata always have the madreporic canal in communica- 

 tion with the surrounding medium by an opening on the dorsal surface, but eventually the 

 canal loses its connection with the exterior so as to hang loosely in the peritoneal cavity 



1 In my memoir on the Elpidia glacialis I also described some large wheels which differ most strikingly in shape 

 from those of other Elasipoda, and present the greatest resemblance to those occurring in the Apoda. Danielssen and 

 Koren are doubtless right in supposing that these wheels had accidentally stuck to the rough surface of the integument, 

 and I feel the more convinced of it as I could never find them in more than one single individual. 



2 Beitrage zur Naturgeschichte der Synapta digitata, ii., Dresden, 1864, pp. 36, 37. 



3 Studien fiber die Entwickelung der Echinodermen und Nemertdnen (Memoires de l'Acad, imp. d. Sc. de St 

 P^tersbourg, vii. si5rie, torn, xiv., Xo. 8, 1869, pp. 6, 7, pi. i. fig. 11). 



(ZOOL. CHALL. EXP. PART XIII. 1881.) N 19 



