MONOGRAPH OF THE EXISTING CRINOIDS. 315 



find a circlet of small free unmetamorphosed iiifrabasals surrounding the central 

 apical plate (fig. 572a pi. 7). 



In the pentacrinites the proximalo never becomes attached to the calyx, but is 

 continually reduplicated, each reduplication as it is formed being shoved away from 

 the calyx by the formation of another between it and the calyx plates, all the multiple 

 proximales later becoming separated from each other by the intercalation of a 

 definite number of so-called nodals (fig. 127, p. 197). Thus there is no opportunity 

 offered for the infrabasals to fuse with the proximale, and so in the pentacrinites 

 we find them forming a definite circlet of minute plates within the circlet of basals 

 and entirely concealed by the column (figs. 566-568, pi. 7). 



In the Plicatocrinidae (figs. 144,p. 207 and 145, p. 209) there is no evidence what- 

 ever of the possession of infrabasals, and also there is no evidence that they ever 

 existed in any of the ancestors of the family, the Plicatocrinidae being as anomalous 

 in this regard as they are in respect to their columns. In all the other recent 

 forms, however, infrabasals are either actually or potentially present. 



Among the recent comatulids, though all are shown to be dicyclic by the 

 application of Wachsmuth and Springer's law, only three species, all belonging to 

 the same family and two to the same genus, are definitely known to possess infra- 

 basals, and in all of these they are present as individual plates only in the very 

 young pentacrinoid, at a very early stage fusing with the topmost columnal or 

 proximale to form, in conjunction with it, the centrodorsal. 



Infrabasals have been conclusively demonstrated in Antedon mediterranea by 

 Bury (figs. 569-571, pi. 7), and in A. adriatica by Seeliger. I have found them to 

 be large and well developed in Promachocrinus Teerguelensis. 



Observations which seem to show that they are not developed in the young 

 have been made on Antedon petasus (Mortensen), A. bifida (Wyville Thomson, W. 

 B. Carpenter, P. H. Carpenter, Perrier, and the present author), A. moroccana 

 (Perrier), Compsometra loveni (the present author), Hathrometra prolixa (Mortensen 

 and the present author), H. sarsii (M. Sars), Ptilometra mutteri (H. L. Clark and 

 the present author), Comactinia meridionalis (Mortensen and the present author), 

 and Comanihus wahlbergii (the present author). 



Most of these observations, however, can not be considered as at all conclusive, 

 as the material available for study was very limited. 



In Atelecrinus balanoides P. H. Carpenter noticed that within the ring formed 

 by the persistent unmetamorphosed basals excessively delicate processes project 

 inward from near the lateral margin of each basal; it is possible that these proc- 

 esses are the remains of infrabasals, which have been for the most part resorbed. 



In Antedon mediterranea Bury found that the infrabasals make their appear- 

 ance in the larva early on the seventh day. They are found at the posterior (i. c., 

 proximal) end of the series of columnars, and in form resemble small basals, though 

 they are developed at a much deeper level and are usually nearer the posterior 

 end of the body than the two ventral basals. They are typically three in number 

 (rarely four or five) and are at first equal in size; but after a while two of them 

 begin to grow more rapidly than the third, eventually becoming about double its 

 size. The smallest infrabasal lies in the anterior radial area of the adult, cor- 



