382 



SCIENCE 



[N. S. Vol. XXV. No. 636 



way aeeonnt for the conditions presented 

 by the higher plants. 



The existence of dioecious races with a 

 widely varying percentage of the sexes 

 is also against the assumption that sex 

 is transmitted as a Mendelian character. 

 Heyer found the hemp plants about Halle 

 showed a ratio of 100 males to 114 females. 

 Fisch at Erlangen found the ratio was 100 

 males to 154 females. 



The conditions noted as to the stage at 

 which sex differentiation occurs, etc., sug- 

 gest at least the universal presence in plant 

 cells of the potentiality for development in 

 the direction of either sex when the neces- 

 sary environmental conditions are given. 



Robert A. Harper 

 Univeksitt of Wisconsin 



SEX-DETERMINING FACTORS IN ANIMALS 



There are few biological questions that 

 appeal more directly to the human race 

 than whether the sex of the child can be 

 determined by the external conditions 

 under which the parents live, or whether 

 the conditions are internal and, therefore, 

 beyond the power of control. This prob- 

 lem has been examined by the statistician, 

 argued by philosophers, discussed by the 

 naturalists and exploited by the quack. 

 Theories of sex determination have flour- 

 ished like weeds, and, while perennial, are 

 apt to be like their prototypes, short lived. 

 The history of these theories is, neverthe- 

 less, full of interest and not without signifi- 

 cance. Even a brief survey will bring out 

 the salient points. 



Aristotle refuted the opinion of Anax- 

 agoras that the male comes from the right 

 side and the female from the left side of 

 the father; that of Empedocles, Avho held 

 that the temperature of the uterus influ- 

 ences the sex of the offspring; and that of 

 Democritus, who suggested that the excess 

 of the male or of the female element is the 

 essential factor. Aristotle, in turn, held 



that the temperature of the germ-material 

 determines the sex, for, he said, more males 

 are born to young and to old parents than 

 to those of middle life, because in youth 

 the temperature of the body has not 

 reached its maximum and in old age it has 

 begun to abate. In recent years we find 

 that one external factor above all others 

 has been supposed to have an intimate rela- 

 tion to sex determination, namely, nutri- 

 tion. An experiment of Landois in 1867 

 furnished the first erroneous evidence in 

 favor of this view. He claimed that he 

 could produce at will males or females of 

 the butterfly Vmiessa by regulating the 

 amount of food. A similar conclusion was 

 reached by Mrs. Treat and by Gentry in 

 1873. Riley, Bessels, Briggs, Andrews, 

 Fletcher, and Kellogg and Bell have shown 

 that no effects of this kind are produced 

 by starving or by feeding. At most there 

 occurs a greater mortality of the female 

 caterpillars through starvation, leaving 

 more males alive. The futility of the ex- 

 periment is now manifest, since it has been 

 shown that the reproductive organs of the 

 male and female are already laid down 

 when the caterpillar leaves the egg. 



Equally inconclusive have been the ex- 

 periments with the tadpole of the frog. 

 The work of Born and of Yung has 

 been upset by the experiments of Pfliiger, 

 Cuenot and of Richard Hertwig. The 

 earlier observers failed to take into account 

 the great mortality of the tadpoles kept 

 under artificial conditions, hence a possible 

 source of error is present in their results, 

 and the conclusions are unsatisfactoiy so 

 long as we do not Imow whether in the frog 

 one sex is more susceptible than the other 

 to unfavorable conditions. Aside from 

 this possibility there seems to be something 

 very peculiar about the proportions of the 

 sexes in these amphibians. 



Diising has applied the statistical meth- 

 od of study to the proportion of males and 



