ix PKOTOSTOMA THEORY 497 



supra-oesophageal ganglia and the suprarectal commissure. According 

 to Sedgwick this stage in the development of Peripatus repeats the 

 features of an Actinozoon-like Coelenterate ancestor, not merely of 

 Peripatus, and therefore of Arthropods in general, but of such other 

 groups as Annelids, Molluscs and Vertebrates. 



It will be noted that on this protostoma hypothesis an important 

 physiological distinction has at an early period of evolution marked 

 off the Vertebrates from the other groups mentioned. This distinc- 

 tion came about with the acquisition of different habits of movement. 

 In the stem which gave rise to Annelids, Arthropods, Molluscs, 

 movement took place with the neural surface next the substratum 

 (as in those modern Medusae which are able to creep on a solid 

 surface — e.g. Cladonema), while in the Vertebrate stem on the 

 other hand the neural surface was directed away from the solid 

 substratum (as in the modern Actinian when it creeps). This differ- 

 ence in the position of the body in relation to the substratum 

 would naturally lead in time to the different types of dorsiventrality 

 so apparent in the fundamental organization of the two diverging 

 stems. It is frequently stated by critics of the protostoma hypothesis 

 that it involves a reversal of dorsal and ventral sides during the 

 evolution of Vertebrates from their invertebrate ancestors but it 

 will be gathered from what has been said that this criticism rests 

 on a misunderstanding. 



It will be readily seen that the protostoma hypothesis success- 

 fully explains the four categories of puzzling facts already enumer- 

 ated. The paired appearance of the gastrular roof would be a 

 reminiscence of the fact that originally it was actually paired : the 

 split along the back of the abnormal embryos would mean the 

 temporary re-appearance of the ancestral split or mouth : the primitive 

 streak would be the scar along which the lips of this ancestral mouth 

 or protostoma underwent fusion : and the converting of blastopore 

 now into mouth now into anus would be an imperfect reminiscence 

 of the fact that in phylogeny it gave rise to both. 



On this hypothesis the various signs of a split along the neural 

 surface of the vertebrate embryo, whether in the form of a dorsal 

 furrow or a primitive streak or an actual opening, are interpretable 

 as reminiscences of the protostoma slit which traversed the neural 

 surface of the Actinozoon-like ancestor. 1 It is of interest to notice 

 that in two Vertebrates at least there exist what seem to be obvious 

 •traces of neural rudiment extending round behind the anal part of 

 the protostoma precisely as in Peripatus. In Fig. 221, B, is shown 

 an embryo of Lepidosiren spread out in one plane, with the neural 

 rudiment in the form of a ridge which is continuous behind the 

 blastopore or anus. If it be reflected that this opening may be 



1 That the primitive streak and primitive groove ar? closely related to the gastrula 

 mouth was perceived by Rauber (1877) but a clear evolutionary explanation of this 

 relationship was first given by the protostoma theory of Sedgwick (1884) and Hertwig 

 (1892). ' „ 



VOL. II Z K 



