8 THE VOYAGE OF H.M.S. CHALLENGEE. 



additional ones, as new loops immediately below the odd terminal tentacle, being identical 

 with that in the Starfish. Several recent writers on Echinoderms deny the existence of 

 this odd terminal tentacle, the homologue of the odd ocular tentacle of the Starfish (see 

 Embryol. Starfish, pi. viii.). I would refer them to the figures of young Arbacia in the 

 Revision of the Echini, p. 735, and of young Echinids {Strong ylocentrotus drobachiensis) 

 in the Mem. Am. Acad., 1864, for figures of this terminal tentacle, and to the Revision of 

 the Echini, pi. x. 1872, part 3, and to the figures of young Goniocidaris canalicidata in 

 this memoir (PI. II.). There is nothing to show that the interambulacral zones in the ear- 

 liest stages at which they can be detected, do not consist at the beginning of several 

 plates, more or less rudimentary, all appearing at the same time. It seems to me more 

 natural to suppose that in the Clypeastroids we have the madreporic body in the neutral 

 position, indicating the mode in which the madreporite passed from an imstable con.dition, 

 owing to the presence of an apical anal system, to a stable one, due to the withdrawal of 

 the anal system to one of the interambulacral areas, which then became the principal 

 guide in fixing the position of an antero-posterior axis until the madreporic body again 

 had a tendency to encroach upon certain parts of the genital system in the Petalosticha, 

 when the position of the axis was again defined by the position of the anus and of the 

 simple ambulacrum. 



Far too much weight has been given to the order of appearance of the plates of the am- 

 bulacral and interambulacral areas in this discussion. The coronal plates, as is well shown 

 in young Echinids, while divided into ambulacral and interambulacral areas, do not, as far 

 as we have been able to trace their appearance, develop in such a regular and fixed man- 

 ner as to enable us to determine the axis of the Echinids from the order of their origin. 



COMPAMSON OF THE CORONAL PlaTES OF THE TeST Uf DIFFERENT FAMILIES. 



Among the Clypeastroids, it is only in the younger stages that the interambulacral 

 plates are connected at the actinostome as in Spatangoids. The actinal ambulacral plates 

 soon increase so fast in width as to drive them apart, and in the older stages of some 

 genera ^ the second row of ambulacral plates forms a continuous ring round the actino- 

 stome, while in others ^ the odd posterior interambulacrum still extends connectedly to 

 the actinostome, as it is in the adult of Echinocyamus and Laganum, and in others form- 

 ing in part a trivium and a bivium as in Rntula, while in Echinarachnius it is the odd 

 posterior interambulacrum of which the actinal plates become first disconnected. 



Echinoneus comes in at once as a marked exception to Loven's theory, as well as all 

 the Clypeastroids, where in no stage do we find that the actinal plates have the 

 characters upon which Loven's theory of an axis is based. 



Lov^n has already called attention to the greater affinity existing between the 



1 Encope, Clypeaster, Arachnoides. ' Mellita: 



