25 



In young specimens the structure of the ambulacra is more easily seen to be 

 typically diadematoid. The structure of the interambulacral areas is well seen 

 from the inside, also on the actinal side. In accordance with the large size of the 

 ambulacral plates the interambulacral ones are very large (high) at the peristome, 

 very short and crowded in the outer part of the granulated portion, and here they 

 seem to split up into several small, irregular pieces at the median end, a feature 

 which is evidently something secondary '). In the other part of the test there is 

 no indication of any division of the interambulacral plates; they are rather distinctly 

 overlapping, the adorai edge being covered (as seen from without). 



The peristome is small, about V4 of the diameter of the test (22 mm. in a 

 specimen of c. 90 mm. diameter oftest); de Meliere finds it even smaller, 11 mm. in 

 a specimen of 60 mm. diameter. The buccal membrane contains numerous irregular, 

 fenestrated plates; the buccal plates do not carry pedicellariæ or spines. The 

 moulhslits are very small and indistinct; the gills are very small, containing the 

 usual irregular spicules. The compact plate in the part of the gills turning towards 

 the test is small and short. 



The spines on the abactinal side and at the ambitus are slender and straight, 

 Va — '/s times as long as the diameter of the test; those just below the ambitus are 

 flattened and a little widened at the point. The actinal spines are short, c. 8 mm. 

 (in DE Meijere's specimen only 5 mm.), curved, very close set, giving the animal a 

 curious, almost bearded appearance. Those at the peristome are a little longer. The 

 milled ring is oblique as in Astropijga, even to an extreme degree in the spines at the 

 ambitus (PI. V. Fig. 10). The spines are vcrticillate as usual in the Diadematidœ; in 

 the small spines the thorns may be bent outwards. In the small abactinal spines 

 the whorls of thorns continue to the very point, in the small actinal ones the end 



') Agassiz has observed a similar splitting up of tlie interambulacral plates in Astropijga 

 (Chall. Echinoidea. PI. X. a. lig. 9.) ,,into irregular!}' shaped independent plates, thus producing inter- 

 ambulacral areas which, as in the l'alæechinidæ proper are composed of more than two vertical rows 

 of plates ... In Astropijga we find that the large interambulacral plates from the edge of the ambitus 

 nearly to the abactinal sv.steni, as far as the external line of primår}- tubercles e.\tends, are made up of 

 two verj' distinct plates, so that in Astropgga as well as in Phorinosoma (viz. Echinosoma tenue, — 

 Chall. Ech. p. 95) we have an interambulacral area, in which the vertical zones are not composed simply 

 of two rows of plates but of four". (Chall. Ech. p. 72). Later on (p. 78) this splitting up of the plates 

 in Astropijga is said to be „limited to a few plates on the actinal side(s) of the test", and on p. 95 it 

 is likewise said to be on the actinal side. Duncan (Op. eit. p. 110) has not been able to see this splitting 

 up of the plates in Astropijga; but whether it usually e.\ists or not, and, if it exists, whether ;t be on 

 the abactinal or actinal side, it is certainU' wrong to see a remnant of the palæeehinoid structure tiierein. 

 If it were really a palæeehinoid remnant, it must evidently be found in the young specimens; but it has 

 not been shown to exist in the young specimens, and it is rather probable that the splitting up of the 

 plates in large specimens may be due simply to the breaking of the delicate plates bj' the handling of the 

 specimens — both in Astropgga, Clm-todiadcma and Echinosoma, and, without doubt, all large forms 

 with such delicate plates. — Unfortunately I have not had sufficient material ai Astropgga for examining 

 the question myself. 



D. K. L) Videiisk. Selsk Skr., 7 R:L*kke, natuivûlensk og niathein AlVi, I 1- 4 



