Origin of the Segmented Worms. 515 



sphincters. Such an anal aperture, acquired of necessity by 

 the lengthening chain of digesting buds, would be earlier and 

 earlier developed by that force in living matter called heredity, 

 which causes it to repeat itself so long as the habits of life 

 demand. 



This specialized aperture, however, would now be in the 

 exact position from which budding had formerly taken place. 

 Although division of highly organized cells is known in the 

 Protozoa, it is certainly commoner in cells not differentiated 

 for special functions. Hence we may suppose that the cells 

 which were specialized as epithelial, muscle-, or nerve-cells 

 to line and regulate the new opening would not divide, at 

 least actively, but the less differentiated zone of cells imme- 

 diately round these would, and by so doing push the specialized 

 anus out in front of them, so that it would always be at the 

 posterior end of the chain. 



At first this anal aperture would have been very simple and 

 small, but as it became more highly specialized the budding 

 zone would retreat from the extreme tip until there appeared 

 a regular " anal segment," the budding zone being always 

 between it and the last-formed bud. This is, in fact, still 

 the law in all animals that develop metamerically — the new 

 segments are added between the anal "segment" and the one 

 in front of it. 



The mechanical difficulty due to the accumulation of faeces 

 being thus simply got over, the possibility of further advance 

 is instantly furnished ; more food would now be passed along, 

 and the posterior segments being better nourished, the budding 

 process at the end could be renewed. The number of buds 

 would, however, still be limited, as, even though faecal matter 

 could now be got rid of, yet as the length increased it is 

 obvious that more faecal matter than food would reach the 

 posterior end of the body, which would thus be badly 

 nourished and cease to grow. This limit could only again be 

 surmounted by the development of a vascular system. 



So far, then, we have arrived at long free-swimming strings 

 of Ccelenterate buds with a mouth at one end and an anus at 

 the other, and with a tendency to produce fresh buds up to a 

 limit between an anal prominence or segment and the last- 

 formed bud. This brings us so near to the typical Annelids 

 that it is worth seeing whether the structural differences 

 which still separate them are insurmountable. Here I might 

 parenthetically remark that it is just such active free- 

 swimming organisms as those we have sketched which attain 

 to higher levels of organization. They, least of all, would 



