June 28, 1907] 



SCIENCE 



1009 



account, in part at least, for the fact that all 

 germ cells are not absolutely pure in hybrids, 

 and it may be thus that in cases of apparently 

 complete return to one parent type, character- 

 istics of the other parent may, nevertheless, 

 crop out from time to time in succeeding gen- 

 erations." 



In the foregoing attempt to reconcile the 

 behavior of chromosomes with the Mendelian 

 principles it is clearly assumed that the chro- 

 mosomes are the mechanism, or at least the 

 chief mechanism of heredity. But turn them 

 and juggle them as we may, we see that it is 

 difficidt to make the conclusions of the cytolo- 

 gist who clings to an exclusively chromosomal 

 theory fit into the facts of heredity, and the 

 difficulty becomes all the greater if it is true 

 that the male and the female do not contribute 

 equally in heredity. The latter difficulty, to- 

 gether with an attempt to explain reversions, 

 led nie to the suggestion, in a former paper, 

 that both cytoplasm and nucleus are involved 

 specifically in inheritance, the cytoplasm of 

 the germ cell representing the more stable and 

 constant form of the animal, and the chro- 

 matin the more individual and variable char- 

 acteristics. If the male contributes chiefly 

 individual qualities, and if we accept these as 

 being borne in the chromosomes, then since 

 these chromosomes are so strikingly paralleled 

 by those of the female in both number and 

 form, we might infer that the chromosomes of 

 the female likewise bear only individual and 

 variable qualities, while the cytoplasm bears 

 the fundamental qualities. 



The material starting points (or inceptors) 

 of definite characters were conceived of as 

 arising through the activity of chromosomal 

 emanations on cytoplasmic materials at dif- 

 ferent stages in the development of the organ- 

 ism. Since the chromatin was looked upon 

 as the more variable cell constituent, it, 

 rather than the cytoplasm, was regarded as 

 conditioning the more variable character- 

 istics, without, however, precluding some 

 nuclear activity at all stages of development. 

 There is every reason to believe that both 



' See ' Spermatogenesis of Normal and of Hybrid 

 Pigeons,' p. 48. 



cytoplasm and nucleus in a given species are 

 distinctive of that species and there is no con- 

 ceivable reason why the cytoplasm any more 

 than the nucleus must be made so anew in 

 each generation. The variable nature of 

 chromosomes is evidenced in the pronounced 

 irregularities which are induced in them 

 through hybridizing or drugging. Further- 

 more, the intimate mingling of the chromatin 

 from the two sexes, the known fact that 

 definite substances (enzymes, etc.) emanate 

 from the nucleus, and the highly unstable 

 nature of nucleo-proteids, would point to the ■ 

 chromosomes as^ at least, an adequate source 

 of variability. Lastly, the facts of differ- 

 entiation, before fertilization, etc., mentioned 

 above, seem to indicate that offspring do not 

 inherit all characters equally from parents 

 but rather only the more superficial ones, and 

 that, therefore, chromosomes, which are de- 

 rived equally from each parent, are connected 

 in some way with such superficial characters. 

 Furthermore, such an interpretation would 

 harmonize better with the fact that the nu- 

 merical variation of chromosomes in closely 

 related groups may be out of all propor- 

 tion to the character differences of the groups. 

 For example. Miss McGill, working on 

 Anax, has found 28 chromosomes, while 

 in some recent studies on a Lihellula I 

 find 23 or at most not over 24 chromosomes. 

 How, then, on the exclusively chromosomal 

 hypothesis of heredity, are we to account 

 for the discrepancy in the number of chromo- 

 somes, such as exists between two members of 

 so restricted a group as the Odonata, which, 

 despite their individual differences, must have 

 the vast majority of their fundamental fea- 

 tures in common? Or how account for the 

 pronounced numerical differences in chromo- 

 somes as recorded by Wilson in different 

 genera of the Heteroptera. For example, 

 Anasa tristis has 21 in the male and 22 in 

 the female, while Proienor helfragei, or Alydus 

 pilosulus, has 13 in the male and 14 in the 

 female. Even related species of the same 

 genus may differ in the number of chromo- 

 somes, and there are indications that this may 

 be true of different individuals of the same 

 species. The numerical differences would 



