Bd. VI: 4) THE ECHINOIDEA. 15 



centre of the plates, as stated in »Revis. of Ech.». A group of small tubercles are 

 found in the middle of the plates; no elevation round the pore. (The very con- 

 spicuous difference in the shape of the ocular plates in A. canaliculata and E. nutrix 

 is well seen on comparing PI. IV Figs. 6, 9 — 10, PI. XIV Fig. 6 with Taf. VI Figs. 

 4, 7 — 8 in Echinoiden d. deutschen Siidpolar-Expedition.) 



The genital plates are somewhat irregular in shape, considerably (about twice) 

 broader than long; the outer edge generally forms a rather prominent point, in 

 which the genital pore is situated. In the male the genital pores are quite small, 

 scarcely larger than the ocular pores; in the female they are large,* occupying the 

 whole point of the plate, and even halfway outside the plate (PI. IV Figs. 9 — 10). 

 In the skin covering the sides of the opening some small, irregular plates may be 

 developed. The madreporite is scarcely larger than the other genital plates. Each 

 plate carries a group of tubercles in the middle, leaving a comparatively broad bare 

 edge, especially along the inner side. One specimen shows a curious anomaly, the 

 genital plate 1 being divided into three triangular plates, two of which are provided 

 with a genital opening (PI. IV Fig. 5). — The anal plates are fairly numerous, small, 

 the outer ones generally not much larger than the inner ones; they are rather thick, 

 each carrying one or a few tubercles, according to their size. — In the younger 

 specimens the anal system is somewhat elevated, in the larger ones the whole apical 

 system is more or less sunken, in accordance with the fact that the species carries 

 its brood upon the abactinal area. (It is, however, to be noticed that also the males, 

 who do not carry the brood, have the apical system sunken.) 



The peristome (PI. IV Fig. 11) is generally somewhat larger than the apical 

 system. The ambulacral plates on the peristome are numerous and close-set; in the 

 larger specimens I find 9 — 11 plates in each series; each plate carries two tubercles 

 inside and one or two outside the pore. The innermost pore remains distinctly 

 larger than the following ones. The inner 1 plates of each two adjoining series do 

 not join, so that the interambulacral plates proceed right to the edge of the mouth; 

 these interambulacral plates are small and irregular, each carrying one or two tu- 

 bercles, according to size. 



The radioles are in all the available specimens short, generally shorter than the 

 horizontal diameter of the test, sometimes, however, surpassing it a little in length; 

 that they may, however, be considerably longer, up to about twice the diameter 

 of the test, is stated by Agassiz, Doderlein and Clark. Even if there has 

 probably been some confusion with other species (thus e. g. it seems very hard to 

 recognize the specimen figured in the »Challenger»-Echinoidea PI. II Fig. 1 as a true 

 A. canaliculata) it can evidently not be doubted that the length of the radioles may 



* Wyv. Thomson, Op. cit. p. 65. says ;>the ovarial openings are extremely minutes ; evidently he 

 has only examined a male in this respect, otherwise he could not have said so. 



