Decembeb 28, 1906.] 



SCIENCE. 



851 



accords with the views of practically all biolo- 

 gists. 



2. The chromosomes retain their identity. 

 This assumption has been arrived at independ- 

 ently of any theory of heredity, and repre- 

 sents the opinion of many investigators. It 

 has not, however, been fully established. In 

 case it is shown that the chromosomes do not 

 retain their identity, the reasoning by which 

 the present theory of heredity is developed 

 may be applied to the subdivisions of chromo- 

 somes which are the ultimate biological units, 

 so that the theory, in its essential outlines, is 

 independent of the theory of chromosome in- 

 dividuality. 



3. Synapsis in organisms exhibiting alter- 

 nate inheritance consists of the union of 

 homologous chromosomes received from dif- 

 ferent parents. This is the view arrived at 

 independently of theories of heredity by the 

 majority of investigators. 



4. If the chromosomes themselves are the 

 ultimate biological units, in the first post- 

 synaptic division the position of bivalent chro- 

 mosomes is so far governed by chance that 

 any given segment of the bivalent is as liable 

 to go to one pole as the other. This conclu- 

 sion was arrived at by Sutton independent of 

 theories of heredity, and its possibilities have 

 been pointed out by Cannon, Boveri, Wilson 

 and others, who have also called the attention 

 of the relation of this phenomenon to the laws 

 of alternate inheritance. 



If independent unit characters reside in the 

 same chromosome, alternate inheritance shows 

 that two homologous chromosomes must ex- 

 change character determinants. There is no 

 cytological evidence of such exchange. It is 

 admitted as a possibility. If it does occur, 

 the theory here developed will apply to char- 

 acter determinants consisting of subordinate 

 constituents of the chromosomes as it is here 

 developed for the chromosome itself. Since 

 there is evidence of the chromosome distribu- 

 tion called for by alternate inheritance and no 

 evidence of such distribution of parts of chro- 

 mosomes, we accept tentatively the simpler 

 assumption that chromosomes retain their 

 identity and are the bearers of hereditary 

 characters. The following facts follow as 



necessary consequences of the above assump- 

 tions : 



Two characters inhering the same chromo- 

 some are transmitted together. Cases of such 

 gametic coupling of characters were cited. 

 The facts of alternate inheritance and of evo- 

 lutionary changes are made clear without re- 

 sort to ids, pangens or other elements subor- 

 dinate to the chromosomes themselves. Illus- 

 trations were given of the mechanism by which 

 Mendelian characters, both simple and com- 

 pound, are distributed to progeny. 



Many mutations may be accounted for as 

 the sudden appearance of characters which 

 have developed through an indefinite period 

 unsuspected until a chance cross revealed their 

 identity. Examples of such characters were 

 given. 



Mendelian unit characters are, for the most 

 part, non-essential characters. When vital 

 characters attempt to become Mendelian, 

 which they continually do, they lead to elim- 

 ination by natural selection. 



The chromosomes of a given nucleus are 

 not essentially differentiated with regard to 

 vital characters. That is, vital characters are 

 functions common to all chromosomes, and 

 hence do not obey Mendel's law, but a different 

 law, which was stated. 



Mendelian unit characters (simple charac- 

 ters) are functions of single chromosomes or 

 a single pair of homologous chromosomes. 

 (Homologous chromosomes are those that 

 unite in synapsis to form a bivalent and are 

 supposed to relate to the same characters.) 

 Homologous chromosomes are of common re- 

 cent descent. 



Non-homologous chromosomes are not of 

 common recent descent, but are of common 

 descent in the distant past. 



Synapsis and reduction require the finest 

 adjustment of function of any office performed 

 by the chromosome, as shown by the sterility 

 of hybrids. Hence, organisms that interbreed 

 freely must vary together if they vary at all. 

 Isolation (cessation to interbreed) is, there- 

 fore, the prime condition in the differentia- 

 tion of species. 



Natural selection affects evolutionary 

 changes in the following ways: 



