AEROPSIDAE. 133 



PALAEOSTOMATIDAE Gregory. 



What to do with the extraordinary sea-urchin described by Gray in 1851 

 as Leskia mirabilis has always been a puzzle, the peristome is so perfectly unique. 

 Lov6n considered it the "type" of a "tribe" which he called Palaeostomata, 

 and Mr. Agassiz, in the Revision, placed it in a family by itself, the Leskiadae. 

 Mr. Agassiz recognized that the name Leskia was preoccupied and he accord- 

 ingly used Palaeostoma instead, but he nevertheless retained the family Leskia- 

 dae. And oddly enough, he referred the genus Palaeostoma to Loven, 1867, 

 p. 432, although Loven had not suggested a generic name and in all his writings 

 of 1867 and 1868 used the generic name Leskia. Loven's Palaeostomata was 

 the name of a tribe and while the generic name is a natural derivative therefrom, 

 it is Mr. Agassiz's name and not Loven's. 



Palaeostoma. 



A. Agassiz, 1872. Rev. Ech., pt. 1, p. 80, 147. 

 Type, Leskia mirabilis Gray, 1851. Ann. Mag. Nat. Hist., (2), 7, p. 134. 



Originally monotypic, this genus remains so today, and no nearly allied 

 Echini have yet been found. The lone species is known only from the East 

 Indian region and as a rule from small specimens. Loven has critically studied 

 the test and de Meijere and Koehler have each contributed to our knowledge 

 of the spines, pedicellariae, etc. The pedicellariae are of the usual three kinds, 

 globiferous, ophicephalous, and small tridentate; but they are not at all dis- 

 tinctive and beyond the fact that they are of a generalized type, they throw 

 no light on the relationships of Palaeostoma. 



AEROPSIDAE, fam. nov. 



It is with no little hesitation that I associate together in a single family, 

 the two remarkable genera Aeropsis and Aceste. They agree in the primitive 

 condition of the peristome and in the possession of the peripetalous fasciole 

 alone, but in other respects each is modified in different ways. It seems neces- 

 sary to unite them in a family by themselves or else to make a family for each 

 and I prefer the former course. My decision is strengthened by Mortensen's 



