BCHINOIDBA. I. q^ 



mine*. The better acquaintance, however, does not grant that Agassiz is right, on the contrary we 

 find that we have here an especially important systematic character. All the genera with deep slits 

 of the test agree also in other respects, as will be shown hereafter, and form a separate, distinctly 

 limited group (that is to say in such a way that not all the forms belonging to this group have deep 

 slits of the test, but that all forms with deep slits of the test belong to this group; for in some small 

 forms no doubt belonging here, the slits of the test are not very large). The group of Les 

 Schizechiniens of Pomel is completely correct — the only correct thing in all the systems 

 hitherto given. 



The form of the test plays a very great part in the previous systems; that all oblong forms 

 belong to the Eehinometridae is considered as a matter of course. Even by Agassiz, who character- 

 izes the family Echinometridce as «having always more than three pairs of pores to each arc , Para- 

 salenia is referred here, although it has only three pairs of pores in each arc; but it is oblong, and 

 accordingly it must be an Echinometrid! That the obliquity, however, is a character insufficient for 

 being the base of a family Echinometridce, has been justly emphasized by Agassiz (Rev. of Ech. p. 436). 

 In Stoniopnenstes there is in large individuals an indication of obliquity, and there are in Echino- 

 metra, in one and the same species, specimens in which the elongation of the axis cannot be traced*. 

 — Alreadv Stewart (381) has called attention to the fact that Parasalenia is distinguished from the 

 Echinomctrider , to which family most would, I should think, refer Parasalenia , in the structure of 

 the spines and the pedicellarise. According to my examinations that quite corroborate the observa- 

 tions of Stewart, there can be no question of referring Parasalenia to the Echinometrids. And so 

 the obliquity of the test must be dropped as a reliable character; not every oblique Echinid can before- 

 hand be taken to be an Echinometrid. That the obliquity is not the same, the morphological axis 

 not being in the same proportion to the longitudinal axis in all the oblique forms, has been shown 

 bv Job. Miiller 1 ), and again emphasized by Bell (op. cit), who according to this fact distinguishes 

 between EchinometrincR and Heterocentrotina. 



As consequentlv none of the characters hitherto used, with the only exception of the slits of 

 the test, have any greater systematic importance, we must seek other characters, by means of which 

 we can set this chaos right. The characters, of which there can be any question, are the following: 

 the structure of the test, the apical area, the spines, the gills, the buccal membrane, the inner ana- 

 tomical structures, especially the dental apparatus and the auriculae, the sphaeridice, the spicules, and 

 the pedicellarias. 



The structure of the test cannot be expected to yield more important characters; if such were 

 to be found they would no doubt have been found long ago, as the attention has hitherto almost 

 exclusively been directed to the form of the test, the arrangement of the tubercles etc. in the descrip- 

 tions. The systematic attempts mentioned above, show to a sufficient degree of how little value the 

 characters found here are. One feature of not quite small importance is found, however, which seems 

 to have been quite overlooked by almost all later authors, viz. that in several forms only every other 

 ambulacral plate has a primary tubercle, while in others every ambulacral plate is provided with such 

 a one. Only in Liitken (op. cit. p. 87) I have found a remark that it is not always the case that 



') Uber den Bau der Echinodermen. Abh. d. Berl. Akad. d. Wiss. 1853. p. 128. 



