Oiiyiu of the Se<jmented Worms. 513 



We merely assume, tlicn, tliat a certain numljer of forms 

 remained free-swimming lon^enou^^h to feed and to pjrow and 

 to put out buds. Now, without ^oing so far as to say that 

 there is only one ])lace where sucli a free-swimming form 

 could bud and yet C(Mitinuc to pursue its active life, viz. at its 

 hinder end, it is certainly clear that the most likely place for 

 a bud to appear would be at the posterior end. Not only 

 would any other j)lacc make forward locomotion impossible, 

 but at the hinder end the conditions seem to bo more favour- 

 able tiian anywhere else. The external pressure would here, 

 in the wake of the animal, be less than anywhere else in the 

 body — indeed, during rapid forward swimming there would 

 be something like suction at this spot. Now if this is true 

 for the first bud, it is surely more true for the second, and it 

 seems clear that we get all the conditions for rapid serial 

 budding : for the addition of the first bud, by increasing the 

 number of available cilia, would propel the animal faster 

 through the water ; this greater speed would enable more food 

 to be swallowed, part of which would find its way into the 

 bud, while at the same time it would still further diminish 

 the external pressure of the water at the extreme hinder end, 

 and we have still more favourable conditions for the forma- 

 tion of a second bud behind that already formed, and so on. 

 If there is any probability in this argument at all, it would 

 seem that the more buds there were the faster would buds be 

 produced, until some limit, presently to be discovered, was 

 reached. 



But apart from this argument, and merely bearing in mind 

 the wealth of colony forniiition developed among the Coeleu- 

 terates, is there not full justification for believing that round 

 the shores of the ancient seas, soon after the primitive forms 

 appeared, there would be free-swimming individuals trailing 

 behind them longer or shorter strings of buds diminishing ia 

 size progressively backwards ? Of these, some, we may fairly 

 assume, would sooner or later again give up more or less com- 

 pletely the laborious free-swimming life and become creeping 

 forms. It is in this direction that I should be inclined to 

 look for the ancestors of some of the other members of the 

 ancient class " Vermes " *. 



• The Tiirbellarii\n Mici-ontomum linearr, Ehr, still swims about trailing 

 short striiifrs of buds aftor it. I was ouco fortunate enoujrii to got a 

 glimpse of it living, under a higli power, and saw very clt-arly what 

 appeared to be typical nematocyj'ts scattered here and there in its skin. 

 The derivation of the Phitiholnunthes from primitive Cre-lenterates, which 

 gave up a free-swimming for a creeping manner of life, has much to be 



