ro4 H. H. Newman 



I. INTRODUCTION 



The experimental work on heredity of the last decade or so 

 has, I believe, dealt too exclusively with definitive characters and 

 has overlooked the origin and development of these characters. 

 The usual method of procedure in such experiments is to inter- 

 breed two species or varieties exhibiting well marked differenti- 

 ating characters, usually of a superficial sort, with the idea of deter- 

 mining for each character of the one species or variety whether it 

 blends with or dominates the corresponding character of the other 

 species or variety. The characters studied are nearly always 

 definitive characters and can be observed only in advanced young 

 or in adults. 



Such work, it seems to me, overemphasizes the importance of 

 one stage in the developmental process. That the definitive 

 stage of a character is a comparatively fixed one does not obviate 

 the necessity of studying the origin and ontogeny of such a charac- 

 ter. But in the kind of work referred to above all other stages 

 in the process, no matter how interesting and instructive, are 

 overlooked. This tendency to regard the condition of characters 

 in the adult as the whole of heredity has led to the publication of 

 this paper as a plea for the study of heredity as a developmental 

 process. In the present work it appears that sometimes the ma- 

 ternal and sometimes the paternal influence is uppermost. An 

 individual or a strain may in early stages show a predominance of 

 maternal characters and later a predominance of paternal, while 

 in between the first and last stages may be seen an apparent struggle 

 between the paternal and maternal forces, expressing itself in an 

 alternating dominance of one or the other. 



The work consists of observations of the comparative devel- 

 opmental histories of two pure breeds and their reciprocal 

 crosses, the study being continued from fertilization until long 

 after hatching. It makes no pretense of being a thorough-going 

 study of Teleost embryology. Only such points in the devel- 

 opmental processes are noted as proved themselves efficient for 

 the comparative study of rates of development, sizes of whole 

 or of parts, degree and kind of pigmentation, functional activity. 



