160 ORTHOGENETIC EVOLUTION IN PIGEONS. 



before. The unit-characters are assumed to preexist in the chromosomes and to 

 stand in need of transportation from the nucleus to the body of the cell in order to 

 develop. Minute particles are known to pass out of the nucleus into the cytoplasm, 

 and hence such migrations of "pangens," if they exist, would not startle us. 



But the whole difficulty still remains. If an innumerable host of specifically 

 distinct unit-characters are let loose in the cell-plasma, how are they to reach pre- 

 cisely predetermined points in the organism, and at just the time when needed? 

 It is right here that the theory breaks down, for the difficulty is not one that further 

 investigation may hope to solve, but one that lands us in hopeless speculation. So 

 long as the primary assumption is that of ready-made unit-characters, specifically 

 distinct and independently variable, whether located in the nucleus or in the cyto- 

 plasm, or in both, the problem of development will remain inscrutable. 



De Vries does not overlook the difficulty, but he consoles himself with the fact 

 that we have still much to learn in regard to the "extremely complex processes of 

 nuclear division," in which, accepting the current view, he believes the purpose 

 to be "manifestly a determinate distribution of hereditary characters to the two 

 daughter cells." "The fact," continues de Vries, "that we do not know how the 

 hypothetic pangens are held together (in a determinate order), is therefore no ob- 

 jection to this assumption, that they are so held in order." 3 



The refuge here becomes precarious when we reflect that heredity can be just 

 as perfect by amitotic as by mitotic division, and that regeneration in Protozoa 

 may take place in pieces containing a whole or even a fragment of a nucleus, without 

 the aid of a distributing mitotic process. Moreover, when we consider the recent 

 experimental demonstrations showing that hereditary processes may run on to a 

 stage as late as the trochosphere in the egg of the annelid Chcetopterus, without a 

 single division of the nucleus, either mitotic or amitotic, it becomes abundantly 

 clear that differentiation and its accurate localization do not depend upon any 

 mitotic distribution of hereditary units. 



I believe that these experiments settle once for all the untenability of all theories 

 of heredity based upon the hypothesis that the nucleus furnishes the regulating 

 machinery for the delivery at definite times and places of appropriate specific heredi- 

 tary units. 



As we review the grounds on which such theories have been erected, we can 

 hardly fail to see that they all find their exigencies in a few gratuitous assumptions, 

 chief of which is that heredity means transmission or transportation of qualities 

 from parent to offspring, or from nucleus to cytoplasm during development. Since 

 Darwin's herculean effort to meet this imaginary exigency through the hypothesis 

 of pangenesis, we have seen the doctrine of "acquired characters" practically dis- 

 posed of, and the "transportation hypothesis" consequently reduced to the "intra- 

 cellular" field. Here the assumption still lingers, in spite of the universally accepted 

 fact that the likeness of sister germ-cells does not depend upon transmission in the 

 strict sense of the word, but upon the fact that each represents a moiety of the 

 mother-cell, in which all essential parts were duplicated before and during its divi- 

 sion as a whole. The mother-cell exists no longer as one cell, but as two cells — each 

 an exact copy of the original, with a full and equal equipment. All this marvelous 

 likeness was generated within the mother-cell, by assimilation, growth, and division. 



' Intracellular Pangenesis, page 67. 



