Origin of the Segmented Worms. 517 



to maintain their efficiency. Further, this growth and ever 

 incrcasinfi; concentration of the nniscuhir systems of the imli- 

 vicinal bnds wouhl necessarily be accoinpanidl by an ever 

 increasing development of their nervous systems. These also 

 would require to be well nourished and cleansed. The 

 feeding and the carrying away of the waste products of this 

 developing neuro-muscular sheath could hardly be carried on 

 in the original Coclenterate wall. Structural modifications, 

 increasing in sj)ccialization as the systems they subserved 

 became more and more perfect, were bound to take place ; 

 these, I think, we may sai'ely outline. In the simpler stages 

 scattered cells with fluid spaces would appear in the mesen- 

 chymal layer; their increase would result in the development 

 of a parenchymatous tissue, with still larger fluid sj^aces, for 

 a compact tissue would be a hindrance to contractility. 

 These spaces would gradually separate the ectoderm from the 

 endoderm — i. e., on the one hand, the ectodermal layer and 

 locoraotory muscles from the digesting endoderm, the only 

 source of nourishment; arid, on the other hand, the digesting 

 endoderm from the ectoderm, the only breathing surface. 

 Further advance along these lines would necessitate definite 

 streams, and ultimately a vascular system, collecting nutri- 

 ment from the digesting layer and oxygen from the ecto- 

 dermal. In this way, then, the endoderm and the ectoderm 

 would be gradually divorced from each other by an increasing 

 development of intervening tissue. The former (endoderm) 

 would be ultimately confined to a central hollow strand, the 

 alimentary canal, suspended and held in place by membranous 

 tissue and wraj)ped round by muscles and cells split off" from 

 the early mesenchymatous or parenchymatous layer already 

 described. In passing we may note that though the deep 

 columnar epithelium characteristic of the alimentary canal of 

 the Annelids could doubtless be developed at any time, still 

 it is just possible that it may be an expression of the shrinkage 

 of the surface due to the detachment of the endoderm from 

 the outer wall and to its becoming the inner lining of a 

 narrowing axial tube. The suggestion arises naturally from 

 the remarkable difference seen between the endoderm of an 

 expanded and that of a contracted polyp. 



Here, then, we have a very simple and natural origin for 

 the mesoderm in the ^letazoa, the activities of our string of 

 buds supplying us with an adequate reason for its develop- 

 ment. The case is, I think, a strong one, for the muscles in 

 each bud necessary to enable it to take its share in giving 

 to the whole string a serpentine movement would have to be 



