18 HAWAIIAN AND OTHER PACIFIC ECHINI. 



subfamily of the Clypeastridae. MacBride (1906, Echinodermata, Cambridge 

 Natural History, 1, p. 549) adopts the same four families but, without any 

 explanation of such an inexcusable move, puts Arachnoides in the Laganidae! 

 Evidently then, except for the genus Arachnoides, all writers are agreed on the 

 primary subdivisions of the clypeastroids. As I consider the arrangement of 

 the auricles of fundamental importance, I believe Gregory's classification is the 

 most natural of those hitherto proposed, but, as I have already pointed out 

 (1911, Ann. Mag. Nat. Hist., ser. 8, 7, p. 593-605), Arachnoides is so different 

 from all Clypeastridae, I prefer to raise Gregory's subfamily Arachnoidinae 

 to full family rank, and thus recognize five families of Clypeastrina. 



There can be little question that the Clypeasteridae are the least specialized 

 forms, the character of the auricles, the structure of the spines, and the pedicel- 

 lariae all giving weight to this conclusion. From such a stock or a nearly related 

 one, Arachnoides has developed, becoming very flat and discoidal, and losing 

 its pedicellariae in connection with its subarenaceous life. The Laganidae and 

 Fibulariidae show by their miliary spines and pedicellariae, as well a? by the 

 structure of the auricles and interambulacra, that they have come from a com- 

 mon stock, for I cannot believe that the simple features of the fibularine ambula- 

 cra are primitive. They seem to show a secondary simplicity. The Scutellidae 

 are of course the most specialized forms, but which genus is to be considered 

 the most extreme I am not prepared to say. Very likely it is Rotula but my 

 material of that genus does not warrant an opinion. 



Key to the Families of Clypeastrina. 



Auricles separate, each placed more or less clearly on the ambulacrum. 



Test rarely discoidal, and usually not flat; anus marginal or inframarginal; genital pores 



Clypeastridae. 



Test very flat, discoidal; anus supramarginal ; genital pores 4 Uiachnoimdak. 



\uricles fused into a single piece situated on the Lnterambulacrum. 



Test seldom discoidal and never with marginal slits or Lunulas; aboral end of each 

 Lnterambulacrum consists of a Bingle large plate adjoining genital: 

 bulacral furrows on oral surface Bhort and indistinct, or wanting. 

 Petals more or less perfect: madreporiC pores numerous; size moderate to 



large ( 1">-I"il) mm. in length) LB. 



Petals reduced and often rudimentary; only one madreporic pore; a 



small, rarely up to 15 mm. in length Fiiui.ap.ii: 



Teal flat and usually discoidal, often with marginal slits or lunules; aboral ei 

 each Lnterambulacrum with the usual pair of small plates adjoining genital: 

 ambulacra! furrows distinct, at least the posterior reaching to the margin S rsuju 



