Referate. 



205 



eggs would give important indications, but these unfortunately have not 

 been developed to a point where the sex can be determined. Reported 

 ca'^es of apparently epigamic modifications of sex he considers inconclusive, 

 and holds that departures fiom the theoretical ratio i :i have no bearing 

 upon the determination of sex. Such deviations in the sex-ratios are 

 probably in the majority of cases at least due to secondary causes acting 

 before fertilization or after fertilization, to eliminate a proportionally larger 

 number, respectively of germcells or of individuals having the one sex, 

 than of those having the other sex ; they do not indicate a change in the 

 sex of any germcell or of any individual. 



Having arrived at these ccyiclusions, he takes up in the second part 

 of his lecture the relation of the inheritance of sex to the inheritance of 

 Mendelian unit-characters, as shown in (a) crosses between hermaphroditic 

 (monoecious) forms and nearly related dioecious forms, and (b) sex-limited 

 characters. The author answers a number of objections which have been 

 raised by various writers regarding the validity of his conclusions from 

 crosses between Biyonia alba and B. diolca. The reviewer does not doubt 

 the correctness of the author's conclusion that the male of B. dioica is 

 heterosametic and the female homogametic, but still believes that the fact 

 that this is a species-cross tends to weaken such a conclusion, not, as the 

 author here assumes because inheritance in species-crosses may be supposed 

 to be fundamentally different from that between more nearly related 

 forms, but because the near incompatibility of the two species favors the 

 thought that whole categories of eggs or of sperms may have failed to 

 function. For instance we might assume that the eggs of B. dioica as well 

 as the sperms are of two types, but that only one of these types is 

 capable of fertilization by the sperms of B. alba. The results would then 

 be exactly those observed. The assumption of a separate determiner for the 

 arrangement of the sex-organs, /. e., for monoeciousness or dioeciousness, is 

 unnecessary and unwarranted by the observed facts. The dominance of 

 the female character of B. dioica over the monoecious character of B. alba 

 is all that it is necessary to assume to explain the occurrence of females 

 alone in the cross B. dioica x alba. The only proper ground for the 

 assumption of a determiner for the distribution of the sex-organs must be 

 found in the recombinations in F2, but this, owing to the unfortunate fact 

 that the Fi hybrids are sterile, is impossible of attainment. 



The classical case of the currant-worm is chosen as the example of 

 sex-limited inheritance. After giving a clear exposition, using the well 

 known furmulae based on the assumption that the female is Ff and the 

 male ff, he develops two other formulations, one proposed by GOLDSCHMIDT 

 in which the female is FF.Mm and the male FF'MJI, and the other with the 

 stated purpose of bringing the case into accord with the presence-and- 



absence hypothesis, the female being FA'At the male simply A'At the A 

 being the unit for some other character with which the F is assumed to 

 be coupled. This last mentioned rather round about method is not necessary 

 to bring the case into harmony with presence and- absence. This can be 

 much more simply done by definition with the use of one of the other 

 formulae. No formulation can meet the facts in the case without assuming 

 some sort of allelomorphism of determiners, and nothing is gained by the 

 assumption that the sex-genes are coupled with other genes which are 

 themselves allelomorphic to each other. The difficulty recognized by the 

 author in assuming the coupling of the grossnlariala-gene, G, with the 

 absence (f) of the sex-gene F, wholly disappears if instead of "coupling" 



