516 Mr. H. M. Bernard on the 



stand still in the midst of the developing life of tlie sea. We 

 have a right to expect that, if in the age of the Coelenterates 

 any such forms existed, profound structural modifications for 

 the perfecting of their activities would take place in them. 



The difficulty, and the only great ditficulty, which lies in 

 the way of the deduction of tlie typical Annelid from the 

 hvpothetical organism described above, lies in the fact, already 

 mentioned, that the former have paired coelomic cavities. 

 The rise of these in the animal kingdom is still one of the 

 great riddles of zoology. Professor Sedgwick's suggested 

 solution, which has been deservedly widely accepted, would, 

 I think, have satisfied me entirely had division of existing 

 segments been the rule for their multiplication. As it is, 

 their known method of production has always seemed to me 

 to point so unmistakably to serial budding, that nothing which 

 has been said on the subject has been able to deter me from 

 watching for clues in that direction. On the other hand, his 

 argument based upon the blastopore is very complete, and, so 

 far as I can see, can only be attacked as coming too much 

 within the field of purely developmental adaptations. I repeat, 

 then, that this paper will have some value if it suggest a simple 

 and natural line of development for these cavities — that is, a 

 line of development without resort to any hypothetical "change 

 of function," which is one of the convenient stocks-in-trade of 

 the puzzled morphologist ; — if it show that in the further 

 development of strings of Coelenterate buds as free-swimming 

 organisms some such mesodermal structures would inevitably 

 arise, fulfilling essentially the same function from tiieir earliest 

 rudiments to their latest specialization. 



We are safe in assuming that the body-cavities arose, as 

 all new structures arise in the animal kingdom, in response 

 to some physiological need. In the present instance we do 

 not require to look far for that need. In the struggle for 

 existence between such chains of Coelenterates it is not likely 

 that the muscles in the walls would be long kept out from 

 assisting in locomotion. A waving of the body is a far more 

 efficient method of propulsion than tiie rowing of cilia. Hence 

 we should expect the wall-muscles in each bud to be gradually 

 specialized into systems in order to bring about definite 

 serpentine movements of the whole string, and, increasing in 

 size and strength, gradually to take on the whole function of 

 propulsion. But muscles require not only to be fed and 

 oxygenated, but also to be relieved of the waste products of 

 their activities. The more they are used the more elaborate 

 must be the nutritive and excretory apparatus to enable them 



