64 SECTIONAL ADDRESSES 



the individual and the evolution of the stock, the specimens selected for 

 developmental studies should be taken from among those w^hich lie upon 

 the mode at each horizon. It will be shown later that if this be not done 

 a curious inversion of the truth may arise. In this connection it is 

 interesting to note that, though Carruthers had not in mind the specific 

 problems we are now discussing, four out of the six specimens whose 

 development he describes fulfilled this condition exactly, and the other 

 two lay only a little to one side of the mode. As to the sequence in time 

 a little doubt attaches only to his middle pair, for from his records it 

 is not quite certain which of the two specimens belongs to the lower and 

 which to the higher horizon. Nevertheless the intermediate position in 

 time of this pair between the other two pairs is beyond dispute. 



It has been necessary to enter into all this detail because, as will be 

 seen later, very great theoretical importance attaches to this material 

 and it is well to know at the outset its precise standard of reliability. It 

 will, I think, be agreed that the standard is a high one. 



On examining the development of the individuals representative of the 

 stages in the phylogeny of the Z. delanoiiei-Z. disjuncta gens it at once 

 becomes obvious that the penultimate stage in the growth of Z. parallela 

 bears a much closer resemblance to the adult of the ancestral species 

 Z. delatwuei {s. sir.) than it does to the adult of Z. parallela. In like 

 manner the penultimate stage in the development of Z. constricta repeats 

 the sum-total of the characteristics which distinguish the adult ancestor 

 Z. parallela, whilst the antepenultimate stage exhibits a similarly close 

 resemblance to the ancestral adult Z. delanouei [s. str.). Here then is 

 an example which fulfils almost, if not quite, perfectly the requirements 

 of the test imposed by Garstang, and proved beyond dispute that specific 

 recapitulation of adult characters does under some circumstances actually 



take place. 



Turning now to the later stages in the evolution of this gens it may 

 be observed that two tendencies, only faintly indicated in the earlier 

 stages, now become more openly manifested. One is the tendency 

 towards the establishment of radial symmetry. This is expressed, feebly 

 in Z. parallela and more clearly in Z. constricta, by the central narrowing 

 and peripheral widening of the fossula. In those later stages which are 

 referred to as Z. disjuncta a second tendency is rapidly expressed in the 

 shortening of the septa, and their withdrawal from the centre, a tendency 

 which in the earlier members of the gens had affected only the cardinal 

 septum. These tendencies are exhibited in progressive degrees of ad- 

 vancement not only in the late life of successive adult stages, but they also 

 pass back into the penultimate and eventually into the antepenultimate 

 developmental stages of the typical and later forms of Z. disjuncta. Thus 

 the principle of specific recapitulation of adult characters holds good also 

 for these two new tendencies. 



In addition to being new, these two tendencies are also out of accord 

 with and may involve a complete reversal and suppression of earlier 

 tendencies. Thus the assumption of radial symmetry implies the 

 disappearance of the tetrameral symmetry, so characteristic of the typical 

 Zaphrentis ; whilst the shortening of the septa is the reverse of the 



