38 BULLETIN "6, UNITED STATES NATIONAL MUSEUM. 



Vorrill included in his Plutonasteiida! throe subfamilies, of which two have 

 just been mentioned. The third, the Plutonasterins, I am unable to separate by 

 any trenchant characters from the Astropcctinidai of Verrill ( = Astropectinina? of 

 Shiden). This was firet brought strongly to my attention when working on the 

 Albatross Hawaiian collections. The genus Patagiaster, according to Verrill's clas- 

 sification would go in the PlutonasteriniE, yet there is not a single character which 

 separates it sharply from undoubted Astropectinidte ; the same is true of PersepJi- 

 onaster, Thrissacanthias, Dipsacaster, Tethyaster, Plutonaster, Londiotaster, and 

 the recently described Ripaster. Of course, even to-day the Astropectinids are 

 supposed to differ from the Archasteridse in lacking an anal aperture. An anal 

 aperture exists in several undoubted Astropectinidse (Lepty chaster, Psilaster, Per- 

 sephoimster, Astropecten, Tritonaster). No more unstable character can be con- 

 jured up to separate Astropecten and its near relatives from Phitonaster, Dytaster, 

 or Dipsacaster. In the fu-st place the absence or presence of an anus is difficult to 

 determine in some species without sectioning. In the second place, there is good 

 evidence that the character is variable within a genus {Astropecten, Psilaster, Lep- 

 tychaster). At any rate, it is proved beyond perad venture that typical Astropec- 

 tinidse may have an anal pore, sometimes of conspicuous size. Why not then let 

 the " aproctuchous " myth take its proper place in history? 



Verrill (1899, p. 199) has shown that the Archasteridse is an untenable group as 

 defined by Sladen. lie has restricted the family to the genus Archaster Miiller 

 and Troschel. With this course I agree. As noted above, we differ on the dispo- 

 sition of Pontaster and the Plutonasterinse. These are placed in the Plutonasteridae 

 by Verrill, the former being made the type of a subfamily. But in Verrill's diag- 

 nosis of the family (1899, p. 210) it is impossible to find any characteristic mentioned 

 which is not shared by Astropectinidse. In other words, the group is poorly defined 

 antl heterogeneous. After eliminating Miinaster and Pontaster, the remainder (that 

 is, the Plutonasterinse) is much more homogeneous, but the genera involved {Plu- 

 tonaster, Dytaster, Lnncliotaster) are connected with Astropecten by a very satisfac- 

 tory series of intergrades. I have been unable even to make two subfamilies in 

 the Astropectinidse. If anyone wishes to try it, numerous possibly available char- 

 acters are furnished in the accompanying table. Here all the genera are listed, 

 and one may see how futile it is to attempt to separate Plutonaster and its allies, 

 either in a family by themselves or as a part of that taxonomical catch-all, the 

 ArchastcridiP. 



All genera here included agree in having pointed tube feet with double ampullae, 

 typical paxillse (either the astropectinoid type or parapaxillse), two series of mar- 

 ginal plates, an intestine, almost always an intestinal ccecum (one exception, 

 Blal-iaster) , and always superambulacral plates. Their negative characters have 

 already been mentioned after the diagnosis of the family. Minor characters of 

 less than family importance are detailed in the accompanying table. So far as 

 possible the order given to the genera is such that the Plutonasterinje of Verrill 

 would begin after Tethyaster, and probalily would include Persephonaster, as this 

 genus has been classified in the Archasteridse, as also has Psilaster « by Ludwig, 

 BlaJciaster by Perricr (1884), and Tethyaster by Ludwig (1897) and Sladen (1889), 



o PbUonaster abysnicola Ludwig 1905 = Psilaster peclinatus (Fisher). 



