814 



THE AMERICAN NATURALIST 



[Vol. XLII 



are borne by different individuals. By far the larger majority 

 of flowering plants have hermaphrodite flowers. Separation of 

 sexual organs on different flowers or different individuals ap- 

 pears here and there as characteristic of entire families, and 

 relatively often as a specific character, or sometimes only as a 

 variation ; and this in the most distantly related groups. 



Correns entertains no doubt that hermaphroditism is the pri- 

 mary type and dioeciousness the derived. He does not regard 

 as conclusive the arguments advanced by Coulter and Cham- 

 berlain that it is impossible to decide which is the primary con- 

 dition since dioeciousness appears in both the quite low and the 

 higher groups of flowering plants, being related in the former to 

 wind-pollination and in the latter connected with insect-pollina- 

 tion. He observes that sex-separation appears also in other 

 groups without any relation to the high or low degree of other 

 characters. The dioecious condition arises in consequence of 

 the physiological or morphological disappearance of one or the 

 other set of members of the hermaphrodite condition. The "de- 

 generation" of a sexual organ here is really nothing else than an 

 arrest at a certain stage in the development of one or the other 

 element in the hermaphrodite flower (Hofmeister ; Goebel), thus 

 producing the monoecious or dioecious conditions. Such change 

 naturally does not proceed without an alteration of the idioplasm 

 of the species involved; the "anlagen complex" of one element 

 must become more or less incapable of development, or latent. 

 As far as we know this process follows independently in both 

 sexes and, in each, internal and external alterations stand in 

 intimate correlation. Male and female flowers of monoecious 

 species are of like high development. The female is not a male 

 arrested at a lower stage of development. The great differences 

 in size and vitality between germ-cells must be regarded as 

 adaptations. 



The main points involved in the problem of this investigation 

 concern the method by which sex is determined and the time 

 when such determination takes place. Since the embryonic and 

 the adult sex-organs contain the anlagen for the characters of 

 both sexes, sex-determination has to do with the question as to 

 which anlage, male or female, shall develop. These conditions 

 demand that the germ-cells have a fixed sex-tendency already 

 before their union at fertilization. Correns emphasizes also that 

 one must not here think of a separation and distribution of 



