IX. Genetic Relations to other Fchinoderms. 1915 399 



the attachment (e.g. Calceocrinidae) a strong secondary bilateralism 

 has been superimposed on the pentamerism, and has generally quite 

 obscured it. Pentamerism could never have arisen under such 

 conditions. 



Another objection to this imaginary ancestor springs from those 

 sanitary principles that have proved so useful in elucidating the 

 morphology of extinct Pelmatozoa (see especially " Caradocian 

 Cystidea from Girvan §§ 232, 583, 591; Trans. Roy. Soc. 

 Edinburgh, vol. xlix, part 2, No. 6, 1913). The guiding idea is 

 that the excreta should be removed as far as possible from the food 

 intake and the respiratory organs, and at the same time in such 

 a position that they will be carried away by currents. Now the 

 imagined primitive Asterozoon contravenes those principles, because 

 mouth, vent, and hydropore were certainly all on the oral face, and 

 yet we are to suppose that face turned downwards towards the 

 attachment. 



Even if the ancestral Asterozoon, as imagined by Messrs. Gemrnill 

 and Ma cB ride, could scarcely have been viable, it does not follow 

 that the Pelmatozoic conception of the ancestor is any more in 

 harmony with known facts and principles. Indeed, it is asserted 

 that the facts of recent Asteroid embryology render it inadmissible. 

 Let us consider some of these " very serious ontogenetic difficulties". 



A reader who accepts the orientation of the rays adopted in 

 Gemmill's paper on Asterias rubens (Phil. Trans., 1914), will speedily 

 meet with stumbling-blocks. 



In the larval stages of that form and of several others, the anus, 

 the stalk, and the hydropore all lie between the horns of thehydrocoel 

 crescent. The hydrocircus is formed by the closure of the horns in 

 that interradius. It is therefore from this point that MacBride and 

 Gemrnill, very reasonably, start their numbering of the hydrocoel 

 pouches. As the first pouch, they take that which lies towards the 

 anterior end of the bilateral larva : if the embryo star be viewed 

 with the mouth upwards, this is the pouch that forms the left horn 

 of the crescent, so that the remaining numbers follow on in a solar 

 direction. Those authors use the lloman numerals I-V, but, for 

 a reason that will appear presently, it will conduce to clearness if 

 for these particular structures we provisionally use the Arabic figures 

 1-5. The rays of the star with which these hydrocoel pouches 

 eventually become associated are similarly numbered I-V by MacBride 

 and Gemrnill, i.e. ra)' I corresponds with pouch 1, II with 2, and so 

 on. Thus, for the adult structure, Dr. Gemrnill arrives at the annexed 

 diagram (Fig. 1), which is copied from that in his memoir, but 

 turned upside-down so as to render it more easily comparable with 

 the arrangement in Pelmatozoa. Here the adult anus lies in inter- 

 radius I/V, which marks the " Asterid plane" of Cuenot. It is 

 not in this interradius but in the adjoining one (I/II, solar in oral, 

 contrasolar in aboral aspect) that the madreporite and stone-canal of 

 the adult are found. The latter interradius also corresponds with 

 the brachiolarian notch, and bears a most important relation to the 

 sagittal mesentery of the larva, and to the epigastric mesentery of 

 the developing star. 



