642 HAEOLD HEATH, 



cleavage would not be able to relate itself intimately to the mesoblast 

 which has already retreated to a considerable extent within the embryo. 



At this point it is impossible to differentiate the effects of pre- 

 cocious segregation from those due to anachronism. There is scarcely 

 a reasonable doubt but that 2di-i and 2di-'^ were originally second- 

 ary trochoblasts and post-trochal cells, but that function has been 

 lost and they now enter the somatic plate. And it becomes impossible 

 also to determine whether 2d2.ti (Figs. 26, 36) ever possessed the 

 function of a post-trochal cell which has latterly been modified by 

 precocious segregation. 



The other case in which I believe an anachronism has taken place 

 occurs in Crepidula. Its more detailed characters have already been 

 considered on page 600, but it may be said however that in Annelids, 

 Umbrella and Ischnochiton the basal cell of the "Molluscan cross" 

 gives rise to a cell which in the latter form becomes a functional 

 accessory trochoblast. This cleavage in all cases occurs after the 

 primary trochoblasts are two in number in each quadrant. In Crepidula 

 however the division of the primary trochoblasts is very slow and 

 were the accessory trochoblasts to form then as in the other forms 

 the development of the head vesicle would be proportionately de- 

 layed. As it is the head vesicle develops rapidly and the division 

 to produce the accessory cells is omitted. Later on two cells arise 

 as progeny of the anterior arm of the cross and degenerating are 

 cast ofl'. By this time the primary trochoblasts have divided and it 

 is possible that these cells represent accessory trochoblasts. 



These two isolated cases atford very little data for a complete 

 understanding of the significance of such cleavages but they serve to 

 emphasize the prospective character of the early cleavage stages where 

 the blastomeres arise with time and space relations admirably adapted 

 for their development into the future organism. And there are the 

 strongest reasons for believing that these early blastomeres under 

 normal conditions not only enter upon constant predetermined careers 

 but pursue them to the end of their existence. Gradual changes may 

 creep in, and certain cells or organs may become modified and in so 

 doing produce changes in other closely correlated cells or organs but 

 the ontogenetic development appears to take place along remarkably 

 stable lines. It is obvious that a proper position is an essential to 

 perfect development but certain facts in the ontogeny of Ischnochiton 

 and numerous other forms lead to the belief that it is not all-impor- 

 tant. For example the second quartette cell in the posterior quadrant 

 or the posterior arms of the Annelid and Molluscan crosses have at 



