Bd. VI: 4) THE ECHINOIDEA. 73 



tinct, though generally less prominent than in the threevalved specimens. The neck 

 s well developed, in the larger specimens shorter; the stalk is simple. Quite small 

 pedicellaria; of this form might equally well be taken to represent the triphyllous 

 form; there is no character distinguishing them definitely from the smaller tri- 

 dentate pedicellariœ (PI. XIX. Figs. 36, 43). 



The colour is stated by PHlLirPi to be white; this, evidently, depends on the 

 preservation. Most of the specimens seen by me are brownish, generally rather 

 dark, both test and spines; a few of them are dark-violet. 



On many of the specimens is found a parasitic or commensalistic Bivalve Mol- 

 lusc, which, as I am informed by Mr. H. LvNGE, belongs to the genus Lepton. 

 The specimen figured in Pi. X. 13 shows several samples of these Molluscs on the 

 abactinal side. They are especially found in the marsupia and near the peristome. 



— It is highly interesting that it proves to be specifically distinct from the species 

 occurring in exactly the same manner on the nearly related Kerguelen-form, Abatiis 

 cordatus, viz. Lepton parasiticiini Dall. The deep sea between Kerguelen and South 

 America thus forms a boundary for both the host and its parasite (commensaliste). 

 On the other hand the Lfptou-species found on A. cavernosus is likewise found on 

 A. Agassizii, which occurs together with, or in the same regions as A. cavernosus. 



— The other parasitic Bivalve Mollusc, Scioberetia australis, described by BERNARD 

 from A. cavernosus was not found on the specimens in hand. 



The specimens of A. cavernosus before me differ so much in the shape of the 

 test that, on comparing the extreme forms, one would hardh' think it possible that 

 they could really belong to the same species. There are, however, so many inter- 

 mediate forms that it seems quite impossible to distinguish them with certainty by 

 the shape of the test, and no characters being found in the pedicellarije or other 

 appendages of the test by which to distinguish them, they must all be regarded as 

 belonging to the same species, affording a ver)' considerable range of variation. 

 The height of the test I have found varying from 58 — 72 %, the breadth from 90 — 

 96 % of the length. In PI. X are figured some of the different forms. One form 

 alone can be regarded as a separate variety, viz. the var. bidens, which is distin- 

 guished by its peculiar globiferous pedicellariœ (PI. XIX. Figs. 32, 39). They differ 

 very conspicuously from those of the typical form, the valves ending in two long 

 teeth, with a short opening or slit below on the inside of the blade. Also the tri- 

 dentate pedicellariœ (PI. XIX. Figs. 35, 42) differ somewhat from those of the typical 

 form in the blade being more rounded. Though the pedicellarisc thus afford very 

 marked distinguishing characters I have been unable to find any such characters in 

 the test or the other structures, and it has thus been thought right to separate this 

 form only as a variety of cavernosus. 



\0 — I(J0133. Schwedische Sïtdpolar-Expcdition igoj — igoj. 



