194 Proceedings of the Boyal Physical Society. 



annelids, arthropods, and molluscs are offshoots of a common 

 triploblastic stem. On the other hand, there are various 

 points in the ontogeny of various vertebrates which, so far, 

 seem explicable only on the hypothesis of a coelenterate-like 

 ancestor, possessing some of the main features of existing 

 Actinozoa, Most important amongst these are the apparent 

 traces of a once present slit-like protostoma or primitive 

 mouth, forming a continuation of the blastopore forwards 

 along the mid line of the medullary plate. Associated with 

 this is the fact that in various vertebrates the blastopore 

 persists as the definitive anus, and that in some, e.g., Lepido- 

 siren, the primitive nerve rudiment is continued posteriorly 

 behind the anus. The only working hypothesis which fits 

 in with these facts is that we are dealing with the vestiges 

 of a condition of things resembling what we find in the 

 modern Actinian, with its slit-like mouth dilated at either 

 end into an inhalent and an exhalent opening, and sur- 

 rounded by a specially condensed part of the general nerve 

 plexus. 



I have shown elsewhere ^ how, in my opinion, the method 

 of mesoderm formation in the higher vertebrates can be con- 

 nected up by a series of links with the method of formation 

 in Amphioxus, where the mesoderm arises in the form of 

 archenteric or coelenteric diverticula, which are, of course, 

 directly comparable with those of the Actinian. 



Considerable confusion of thought has been introduced 

 into vertebrate morphology by the loose way in which the 

 word " outgrowth " is used ; the optic rudiment is frequently 

 described as an "outgrowth" from the thalamencephalon, the 

 mesodermal rudiments as " outgrowths " from the gut- wall. 

 As a matter of fact, it would be more accurate to describe 

 these structures as being formed rather by being tucked off 

 from the neighbouring parts of the brain or gut-wall. This 

 is important in connection with the question of the exist- 

 ence from the beginning of the nervous connection between 

 central nervous rudiment and myotome, for we see that the 

 nervous bridge in the earliest stage at which I was able to 

 determine its existence in Zepidosiren, connects two points 



^ Quart. Jour. Micros. ScL, vol. 45, p. 3i. 



