474 H. H. NEWMAN 



optic vesicles. Numerous other cases of this sort are found in 

 Moenkhaus's paper, while other crosses between these same pairs 

 of orders give a much lower degree of success. Even Moenkhaus's 

 own data, therefore, fail to support his claim as to the relation 

 between kinship and developmental success. 



d. General considerations concerning the influence of foreign sperm 

 on the development and inheritance of teleost hybrids 



After working, in this field for a number of years, it now seems 

 to me that we are dealing with two quite distinct types of phe- 

 nomena, the first a matter primarily biochemical in character 

 and the second a matter of genetics. When we are dealing with 

 rates of development we are invading a field more closely related 

 to chemistry than to heredity, but when we attend to the reap- 

 pearance in offspring of the specific embryonic or larval charac- 

 ters of the parent species we are deahng with legitimate problems 

 in heredity. In thus excluding from the realm of hereditary 

 effects of foreign sperm those which have to do with mere rate 

 of cleavage and later development, I realize that I am taking a 

 position totally at variance with that to which I adhered with 

 considerable tenacity for several years and which I maintained 

 in earlier papers dealing with Fundulus hybrids. At the time 

 when these papers were written the observations that most im- 

 pressed me were those that had to do with the rate of development 

 of hybrid embryos and the time of appearance of the various 

 embryonic structures. In both reciprocal crosses between F. 

 heteroclitus and F. majalis it happened that the rate of develop- 

 ment of the hybrids was intermediate between that of the two 

 parent species. This was true of the. cleavage rate, rate of germ- 

 ring formation, time of appearance of first pigment and time of 

 establishment of heart-beat and circulation. On this basis the 

 conclusion was reached that, with regard to the great majority 

 of characters studied, the blended rather than the alternative 

 type of inheritance prevailed. Now, however, that I have many 

 cases before me in which the rate of development of the hybrids 

 bears no relation to that of the parents, I am convinced that 



