720 



NATURE 



[February 3, 192 1 



thing like 50,000 specimens. Obscure as the 

 meaning of the phenomena still is, there can be 

 no question that when the full interpretation is 

 unravelled the work will be admitted to have an 

 importance at least proportionate to the astonish- 

 ing labour which has gone to its production. 



In outline the main result claimed is that the 

 various races can be arranged in a scale ranging 

 from the "strongest " to the "weakest," and, this 

 series once established, the consequences of mat- 

 ings made between races occupying different posi- 

 tions on the scale can be predicted with consider- 

 able accuracy. Intersexual females appeared when- 

 ever the male of a "stronger" race was mated 

 with the female of a "weaker." The intersexu- 

 ality in its several degrees might affect all the 

 sexual characters, primary or secondary, and in 

 its higher manifestations the instincts also. 

 Where such a diversity of features is concerned, a 

 quantitative scale must obviously be largely a 

 matter of individual judgment, but it is claimed 

 that the amount to which these females were 

 modified in the male direction was roughly pro- 

 portional to the interval between the parent races 

 on the scale of strength ; and in the extreme case, 

 when the strongest male was mated with the 

 weakest female, the brood generally consisted of 

 males only, which are interpreted as being in part 

 aboriginal, genetically determined males, and in 

 part individuals which would have been females 

 but for the disturbing influence which has trans- 

 formed them into males. 



Other matings led to the production of inter- 

 sexual males. The discrimination between the two 

 kinds of intersexes was not, to judge from the 

 illustrations, so difficult as' one would have ex- 

 pected. The intersexual males appeared with 

 some regularity in Fj from the cross mentioned 

 above (Japanese 9 x European d) as giving 

 all normals in Fj and in certain other 

 families besides. There were also some con- 

 siderable families all-female. Throughout the 

 complicated series of matings glimpses of order 

 appear which suggest that a comprehensive solu- 

 tion is not very far off. It has, nevertheless, not 

 yet been attained. One of the most curious 

 features, as yet inexplicable, is the fact that in 

 the matings giving all-male families females 

 occasionally appear which are perfectly normal, 

 though their sisters are supposed to have been 

 wholly transformed into males. 



The interpretation which Prof. Goldschmidt pro- 

 poses cannot be adequately expressed in a brief 

 statement. He is under the influence of the theory 

 that each sex contains the potentialities of the 

 other, a conception to which it is now not easy 

 NO. 2675, VOL. 106I 



to attach a precise, still less a factorial, meaning. 

 He is disposed to regard the sex ultimately 

 assumed by a given zygote as decided by a 

 struggle or reaction taking place between two 

 components : (i) the sex-factors brought in by 

 X-chromosomes, and (2) a substratum conceived 

 of as inherent probably in the cytoplasm, and 

 capable by its own development of conferring 

 potentialities opposite to those borne by the 

 factors proper. To these opposing elements 

 numerical values are assigned, arbitrarily as it 

 appears to me, and I have been unable to discover 

 in what way the analysis thus offered differs from 

 a restatement of the empirically observed facts, 

 nor is the representation of the all-male and all- 

 female families as alternative end-products of a 

 balanced reaction at all satisfactory. During 

 the period covered by Prof. Goldschmidt 's ex- 

 periments, phenomena closely analogous have 

 been discovered by J. W. H. Harrison in the Bis- 

 toninae. Evidently we are thus brought into 

 touch with a set of facts, probably abundant in 

 nature, which must be accounted for before the 

 problem of sex-determination is disposed of; but, 

 paradoxical as these occurrences are, they do not 

 justify a return to earlier stages of confusion. 

 The problem created by the existence of inter- 

 sexes, gynandromorphs, and other sex-monstrosi- 

 ties has always been realised. The case of the 

 free-martin, though its true nature is now settled 

 by the brilliant work of Lillie (well summarised 

 in Prof. Goldschmidt 's book), proves that influ- 

 ences as yet little understood may be taking part 

 in these determinations. 



An interesting attempt was lately made by 

 Morgan and Bridges to apply the chromosome 

 theory rigorously to a number of mosaic gynan- 

 dromorphs which have appeared from time to time 

 in the pedigreed work on Drosophila. The 

 parental composition being known, it could be 

 shown from the distribution of the sex-linked 

 factors that in nearly every case these curious 

 patchworks might be represented as resulting 

 from a presumably accidental elimination of a sex- 

 chromosome from the affected parts of the body. 

 The result was certainly a striking one ; but this 

 interpretation is not readily applicable to inter- 

 sexual forms which are not mosaics. Admitting, 

 however, that in mosaics something may 

 have been eliminated from the affected 

 patches, the suggestion that this something is the 

 sex-chromosome raises the questions : Why do not 

 the miscellaneous variations, to which the chromo- 

 somes of somatic tissues are conspicuously liable, 

 more frequently show their consequences as 

 somatic patchworks? and, conversely, ■\\'hy are 



