40 THE CYSTJDEA 



biological phenomena of homoplasy and convergence, can also be 

 studied in this class. 



The Cystidea were first separated from other Echinoderms, 

 under that name, by L. v. Buch in 1844 and 1845. His definition 

 laid stress on the fixed condition, the irregularity of the thecal 

 plates, and the absence of arms like those of Crinoidea. Sub- 

 sequent discoveries of stemless cystids, of cystids with radial 

 symmetry in the theca, and of arm-like structures in most cystids, 

 have made the letter of this definition untenable ; but its spirit 

 holds good. The difficulty that this class has presented to 

 systematists is chiefly due to these factors : (1) The rarity and ill- 

 preservation of these old Palaeozoic fossils ; (2) the ancestral 

 nature of the group and the consequent existence of links between 

 it and other groups ; (3) the wrongful ascription to the Cystidea 

 of various genera (e.g. Porocrinus, Stephanocrinus, Hypocrimts, 

 Echinocystis) ; (4) the extraordinary diversity of structure in the 

 class, a feature common to most groups at their origin, and pro- 

 ductive in this case of many lines of development, only a few of 

 which have become so severed from the rest as to be regarded as 

 independent classes (e.g. Blastoidea and Crinoidea, distinguished 

 by all ; Edrioasteroidea, distinguished by a few ; Anomalocystidae, 

 not distinguished, but quite as separate) ; (5) the rapid develop- 

 ment of the class, from the exceedingly simple Aristocystis to such 

 highly specialised forms as Lepadocrinus, Caryocrinus, and Mesocystis. 

 Hence the diagnosis cannot be elaborate, and must be mainly 

 negative. 



Most of the classifications hitherto proposed have been based 

 upon one set of characters; thus Zittel's (1879) adaptation from 

 Johannes Miiller (1854) is according to the structure of the thecal 

 plates (Aporitidae, Diploporitidae, Rhombiferi) ; Barrande's 

 division (1887), not intended as taxonomic, is according to the 

 number of openings in the theca. A far better arrangement is 

 that initiated by Pictet (1857), extended by Bronn (1859), and 

 modified by Dujardin and Hupe (1862); this, however, is rather 

 a key to genera than a classification into orders and families. 

 Attempts have also been made (e.g. Forbes, 184K ; and.Neumayr, 

 1889) to determine the lines leading from the Cystidea to other 

 classes; and on such principles Steinmann (1888) founded his 

 classification into Eucystoidea, Cystechinoidea, Cystasteroidea, and 

 Cystocrinoidea. A classification on true phylogenetic principles 

 was first published by Haeckel (1896), who only failed from want 

 of acquaintance with the facts of Cystid structure. The classifi- 

 cation in this text-book attempts to express the 'actual lines of 

 descent as inferred from an independent study of the fossils. 



The main lines of descent are these. The starting-point is a 

 simple, many-plated, sac-like form (e.g. Aristocystis, Fig. II. p. 44), 



