ANEMONES 4O3 



if it is objected to on the ground that the process of fission has not been 

 observed from beginning to end, or that the web of tissue between the 

 oesophagi is a true perfect mesentery. Possibly my view will not 

 hold against either of these objections ; yet the suggestiveness of the 

 facts cannot be denied. 



In the second place, there is a much more convincing consideration 

 to be offered. We have seen that either one or two siphonoglyphs 

 may occur in non-sexually produced individuals. We have seen also 

 that hexamerism is not dependent on mode of reproduction. If, then, 

 both hexamerism and the diglyphic condition may result separately 

 from a non-sexual process, it follows, from the law of probabilities, 

 that at some time in all probability the two will be associated. In that 

 case, one in five of these diglyphic hexamerous polyps would be regu- 

 lar, according to the same law ; since in a diglyphic hexamerous polyp 

 one siphonoglyph and its pair of directives can occupy but five pos- 

 sible positions relative to one another, and one of the five must divide 

 the non-directives into two equal groups of two each. It may be 

 objected that this mathematical probability accounts for only a very 

 few regular, diglyphic, hexamerous polyps. The number produced 

 non-sexually is not known, it is true. But whatever it may be, non- 

 sexual processes are accountable for it — and that is the first point to be 

 noted. 



If we turn to the general grounds which, in the third place, may 

 be offered in support of the non-sexual origin of regular diglyphic 

 hexamerous polvps, the objection in the last paragraph may be per- 

 ceptibly weakened. 



It is well known that buds tend to develop according to the parental 

 type, among many groups of animals. The buds of Hydra are indis- 

 tinguishable from the individuals upon which they grow. The buds of 

 Peropkora are identical with the budding individuals in all adult feat- 

 ures, notwithstanding that some organs of the buds may not be devel- 

 oped from the same germ layers as are the corresponding organs in the 

 sexually produced form. It is true that among the Coelenterata, such 

 forms as the hydroids exhibit a marked difference between the sexual 

 bud and the individual which may produce it. Among the Ascidians 

 also this same dimorphism is conspicuous. In these cases, however, 

 dimorphism is associated with a division of labor which often involves 

 an enormous change of function and niay also be correlated with differ- 

 ences of environment. 



In Metriditi7n there is not the slightest reason for presupposing any 

 dimorphism on any of these grounds. The species is not colonial and 



