I \I.\M<M|;|M s DltiMKD.K. 85 



strata, and none, at leas! among the Starfishes and Ophiurans, which would 

 .sluiw anv closer relationship to Crinoids than we find in our own time. 

 From our present knowledge of palaeontology, we certainlj cannot assert 

 that the Crinoids arc the direct ancestors of the Starfishes, Ophiurans, or 

 Echini, — leaving the relation of the Holothurians to them out of the 

 question. 



As far as the Echinoids are concerned, from the time of their appear- 

 ance in the Lower Silurian they have constituted as distinct a class as 

 they do to-day. We must look for the origin of the Echinoderms to Pre- 

 Silurian times of which we have no record, unless a more careful stud\ 

 of the Cystideana may reveal among them affinities to the Echini and 

 tin' Starfishes which are not suspected as yet. The Cystideana arc the 

 only group of Echinoderms which date far enough hack to have preserved 

 perhaps some of the transition types from which might have come the 

 Starfishes, Ophiurans. and Sea-urchins. 



Neumayer has well stated that there is nowhere in the younger forma- 

 tions a single transition form, and the appearance of such forms would 

 go far to speak against a theory of descent, while all the transition forms 

 we know are of such characters that we find no difficulty, as I have 

 already noted for the Echini,* in tracing a quasi lineal succession. 



In tracing the homologies among the order of Echinoderms, we are 

 perhaps only comparing structures which have developed entirely inde- 

 pendently in each of the orders of Echinoderms, and which, as Semon 

 remarks, may he due to the similarity of the conditions imposed upon the 

 early types by the modification of the ambulacral systems, and to the 

 conditions imposed upon the development of the skeleton by the radial 

 plan of growth. 



The great irregularity we trace in the radial development of the 

 plates of the few Cystideans we know may be a hint of the mode 

 of development which has been followed independently in the other 

 orders.! To argue from this, as has been done by Semon, that thus far 



* Palaeontological and Embryological Development. An Address before the Am. Assoc, for Adv. 

 of Science, August, 18S0. 



f It seems more natural to look upon the great difference and variability in the development of the 

 calyx plates in the different orders of Echinoderms as depending to a great extent, as has been sug- 

 gested by Carpenter (see Challenger Report, p. 1.72), on their relation to internal organs. 



Wyville Thomson (On the Embryology of Antedon rosaceus. Trans. Royal Society. 186-3) looked upon 

 the modifications of the skeleton which characterize the principal divisions of Echinoderms as depending 

 mainly upon the relative development or suppression of the radial and perisomic systems of plates. 



