94 PROF. W. GAESTANG ON THE THEORY OF RECAPITULATION : 
In selecting this type from the few that fulfil the necessai'y conditions, I have 
naturally not overlooked the fact that in the two most doughty advocates of 
Haeckel's law in this countrj', we also possess two of the foremost experts in 
Echinoderm embryology and Crinoid palaeontology respectively. If I err in 
my selection, or statement, of facts to be brought into prominence, they will 
know, I think, that it is from inadvertence and not from intention. The 
test of palfeontology cannot, of course, be brought to bear on the origin of 
the stalked condition, or at present, at any rate, of the primal torsion of 
the internal organs of Echinodermata ; but I regard it as established hj the 
form-sequences wliich Bather and his colleagues have traced, that radial 
symmetry was imposed upon the skeleton of an original pear- or sac-shaped 
body by the extension of superficial food-grooves leading to ihe mouth from 
food-collecting tentacles — a view which I understand is shared by MacBride 
(1911, p. 248). The hypothesis that Oystoid, Blastoid, and Clrinoid were 
successive and independent offshoots from an unknown stock that lacked a 
skeleton seems to me to involve the negation of precise morphological evidence. 
Accepting as my basis Bather's masterly sketch (1900) of the jjhyletic 
classification of these groups, and bearing in mind his own cautions (I. c. p. 138), 
as well as the slenderness of the geological record of Permian and Triassic 
forms, the main outlines of the adult ancestry of Antedon cannot have 
deviated far, I think, from the following sequence (the Roman numerals in 
brackets refer to certain figures of special significance in Bather's work) : — 
{?) Prse-Canibrian. — Pra3-brachiate ancestors, first Cystoid, with numerous 
irregular thecal plates, then reduced and approximating to Blastoid 
regularity. Finally an immediate ancestor of Oysto-Blastoid structure 
exhibiting an ill-defined separation between calyx and stalk (c/. 
Gystoidea, viii., xviii.), but with fixed pentameral symmetry and com- 
position of the firm cup, as in Blastoids*, from the Basals upwards 
[Stephanomnus , ii.). From such an ancestor, after development of 
arms, the Monocyclica and Dicyclica, distinguished at first only by 
the exclusion, or inclusion, of Infra-Basals in the cup {i. e. the position 
of the growth-zone), diverged. I pursue further only the Dicyclic 
series, and neglect the Camerata.- 
Cambrian. — The primitive Inadunate : 5 simple arms, distinct from the 
cup ; disk firmly plated with 5 Deltoids (Orals), supporting ambulacra 
above their conjoined edges (cf. Hyhocrinus, xxxvi.). Slightly 
modified, this type survived among Ordovician Cyathocrinoids 
{Porocrhms, Ixxxvi.) : Anal plates (X & RA) present in the circlet of 
Radials ; Posterior Oral, the only madreporite. 
* The stereotyped nionooyolic constitution, sharply separated stalk, and late geological 
■development of Blastoids suggest that this group may be ctmiposed, in reality, of ixedo- 
(/enetic Crinoids, and it would he worth while to extend this hypothesis to some of the 
Cystids themselves. 
