366 



B. C. PUNNETT. 



immense periods of time to a selective death-rate in a uniform environment. These considera- 

 tions coupled with the small number of existing species give us some reason for doubting 

 whether Natural Selection plays much part in determining the number of segments of which 

 these animals are composed. 



Supposing then that Natural Selection is not at work on the meristic characters of the 

 Cephalochorda, there may yet be some other factor active in producing modifications. Sexual 

 selection is clearly out of the question and external fertilization seems to preclude the possi- 

 bility of physiological selection. There yet remains to be considered the factor suggested by 

 Karl Pearson to which he has applied the term Genetic (or Reproductive) Selection. Pearson 

 has suggested that certain characters may be associated with increased fertility. Such characters 

 would tend to assert themselves under a high death-rate, which at the same time need not be 

 selective. To quote his own words — -"Those who deny natural selection must, if we can show 

 the existence of genetic selection, still admit that evolution is a reality, for genetic selection 

 produces a continuous change of type which can only be held in check by more or less 

 stringent selection^" We have then to enquire whether fertility is really correlated with any 

 other characters, — in this case with the number of somites. And in this connection the 

 Cephalochorda are most favourable objects for study not only from their mode of breeding, but 

 also from the fact that the number of gonads affords us a quick and easy method of esti- 

 mating the relative fertility in different individuals". Their numerical range of variation is 

 exceedingly large (cf. Table II.) and their variability as measured by the coefficient of variation 

 (C.V. = 6'4) is extremely high compared with that for the total number of segments. That 

 such gonidial variations are to some extent associated ^vith meristic changes in the total 

 number of somites may be seen in the following table for Heteropleuron maldivense. 



TABLE IV. 



This shews that increase in the number of gonads is associated with a decrease in the 

 number of segments and that the differences are well marked. It may be further shewn 

 that there exists no correlation between the number of gonads and the total length of the 

 specimens (Table V.), or again between the total length and the number of somites in the 

 different specimens. 



Hence the correlation shewn in Table IV. is probably direct and not due to gonidial 

 and .myotomic variations being both dependent upon some other quality such as length. Un- 

 fortunately the data here available are not more than sufficient to warrant these results being 

 regarded as more than a plausible suggestion. Before it would be possible to prove that 

 genetic, and not natural selection, is the factor at work on the meristic segmentation of the 



' The Grammar of Science, 2nd edition, 1900, p. 439. and with a shorter series of them. Tliere seems at present 



^ It is assumed here that the average size of the separate no reason against assuming this but direct observations 

 gonads is approximately the same in individuals with a longer would of course be necessary. 



