178 THE VOYAGE OF H.M.S. CHALLENGER. 



many species of this family, as has been noticed by Troschel, who, however, considered the 

 presence or absence of any fascicle as of considerable specific and even generic importance. 

 It is evident from an examination of many .specimens of this species, that we may have 

 remnants of the lateral and anal fascicle irregularly scattered round the anal extremity 

 either as imperfect anal fescioles or as branches of the peripetalous fascicle or as indistinct 

 subanal and anal fascicles or remnants of the lateral fascicle. 



In a young specimen measuring scarcely an eighth of an inch, and in younger stages 

 (PI. XX.* fig. 9), the anal system is placed within the peripetalous fascicle, so that the 

 second or normal stage, as we have it in the adult, is due to the gradual passage of the 

 anal system from this abactinal position to one below the peripetalous fascicle, and the 

 formation of a new peripetalous fascicle inside of the anal system, and thus at one time 

 the anal system was included within a triang-ular space formed l^y a branch of the 

 original peripetalous fascicle, and the new base of the same turning across the odd 

 interambulacrum between the anal system and the abactinal system. 



These branches are sometimes persistent, and have been also noticed by Troschel, 

 who, however, was not aware of their origin in Faorina. This secondary subanal fascicle 

 usually disappears with age, and is not identical with the permanent subanal or anal 

 fascicle which is formed at a later stage, while the other branch if persistent would form 

 a fascicle above the anal system. The great variation existing in the extent and distinct- 

 ness of the anal fascicle is well shown in the diiferences found to occur in specimens 

 of various sizes from cue and the same locality in Hemiaster cavernosus. It is also 

 plain that the anal fascicle, as such, derived its origin from the peripetalous fascicle, 

 while the subanal fascicle is formed independently, and may exist where no peripetalous 

 fiisciole is found, as in Spatangus, Maretia, and the like ; and many older genera, such 

 as PalcBOtroims, Argopatagus, Pourtalesia, and Urechimis. 



The following description of the manner in which the young are carried in the 

 marsupium formed by the deeply-sunken lateral ambulacra is taken from Thomson's 

 account in the Voyage of the Challenger, vol. ii. p. 231 : — "In the female, the pore-plates 

 of the paired ambulacra are greatly expanded and lengthened and thinned out, and 

 depressed so as to form four deep, thin-walled, oval cups sinking into and encroaching 

 upon the cavity of the test, and forming very efficient protective marsupia (PL XX." fig. 

 6 = fig. 44). The cvarial openings are, of course, opposite the interradial areas, but the 

 spines are so arranged that a kind of covered passage leads from the opening into the 

 marsupium, and along this passage the eggs, which are remarkably large, upwards of a 

 millimetre in diameter when they leave the ovary, are passed; and are arranged very 

 regularly in rows on the floor of the pouch, each egg being kept in its place by two or 

 three short spines which bend over it (PI. XX.'' fig. 2 = fig. 46). Among the very many ex- 

 amples of this Hemiaster which we dredged in Accessible Bay, and afterwards in Cascade 

 Harbour, Kerguelen, there were young in all stages in the breeding-pouches, and although 



