^ SYSTEMATIC SECTIONS. 67 



say that the modern type of Phanerozonia is unknown before Devonic time. It 

 should be added here that complete superposition of the supra- upon the infra- 

 marginals took place in more than one stock and at different times." It is my 

 own belief that the infero-marginalia preceded the supero-marginalia in develop- 

 ment and are of more general importance than these latter ossicles. This point 

 will be returned to later. 



(2) Bather (90, p. 317) points out that I state (Introductory Section to this 

 Monograph, p. 18) that in the older Asterozoa the ambulacralia form a complete 

 floor to the ambulacral groove, and there are no podial pores to allow the ampullae 

 to penetrate into the body-cavity, and adds that " were that view correct, Edrio- 

 uster itself would be more advanced in this respect than the older Palseozoic 

 Asterozoa, and would find its analogue in such a form as the Lower Devonian 

 Xenaster, where the pores have the same relative size and position. ... If 

 pores were absent in all the pre-Devonian Asterozoa, it would be very difficult to 

 understand how tbe relatively narrow pores of Xeuaster were formed. Starting 

 with pressure of an incipient ampulla outside the floor-plates, one would expect to 

 observe a gradual deepening of the excavation until it broke through into the 

 thecal cavity as a relatively wide hole. Such is, in fact, the appearance presented 

 by Prof. Jaekel's drawing of Silaraster perfectus, from the uppermost Ordovician 

 of Bohemia. As our knowledge increases it may be that we shall find among the 

 early Asterozoa, as among the Edrioasteroidea, some genera with podial pores, 

 others without, forming parallel lines of descent. The presence of endothecal 

 ampullae is necessarily dependent on tbe existence in the rays of a thecal cavity 

 large enough to contain them." 



My further studies enable me to amend my previous statement. Careful 

 modelling of the ambulacral groove of Schnchertla, n. sp., has convinced me that 

 podial pores may be present in many primitive Asterozoa between the bases of 

 the adambulacralia and just at the outer extremity of the lines of junction of two 

 neighbouring ambulacralia. In further evolution, this pore gradually eats inwards 

 so that eventually it approaches well towards the middle of the groove. The 

 beginnings of this encroachment can be seen in Xenaster and even in some 

 Ordovician Asteroidea, e. g. in the mouth-region of Proiiiopalseaster clizce (PI. IV, 

 fig. 3). It is obvious that these podial pores are analogous to rather than 

 homologous with the pores of Edrloaster. 



(3) Generally the adambulacralia are equal in number to the aml)ulacralia. In 

 very advanced species of the Promopalasasteridae such as P. magnificns, there is, 

 however, a tendency for the adambulacralia to break up and to lose their form, so 

 that they are irregularly situate to, and outnumber the ambulacralia. These 

 changes appear to be comparable with the changes of the apical plates already 

 dealt with. 



(4) The most primitive arrangement of the adambulacralia with respect to the 



