378 W. Batesoll. 



neg-ative characters as due to the presence of some inhibiting- in- 

 fluence, but in our present state of knowledge there is no certain 

 warrant for such an interpretation. As will be evident, especially' 

 in cases where colour is concerned, a vast number of the observed 

 facts can be readily represented in terms of the presence and ab- 

 sence of definite factors, and it is quite conceivable that Mendelian 

 inheritance in g-eneral may be reducible to such terms; but in view 

 of the difficulties indicated, such expressions must not be adopted 

 with undue haste. 



Recombinations and the production of novelties. 



Mendel showed that the various characters studied by him in 

 Pisiim were separately transmitted. Taking" the distinctive features of 

 the plants and the seeds he determined seven such distinct characters, 

 each depending on a distinct unit-character with its appropriate 

 allelomorph. As the distribution of each pair of allelomorphs is here 

 independent of that of any other pair, the characters belonging- to 

 the several pairs can be recombined in any manner. ^loreover. in 

 such cases it makes no difference how the various characters are 

 compounded in the respective parents. The result for all subsequent 

 generations is the same whether the father is tall, purple-flowered, 

 and yellow- seeded, and the mother dwarf, white-flowered and green- 

 seeded, or whether these features were distributed betAveen the two 

 parents in any other way. The various possible combinations Avill in any 

 case appear in their appi'opriate numerical proportions among the descen- 

 dants. It follows, as Mendel showed, that where two pairs of characters 

 are concerned, of which one in each pair is a dominant, that the 

 average ratio in F.^ will be 9:3:3:1. where the 9 members exhibit 

 both dominant characters, each of the groups of 3 have one dominant 

 character, while the 1 has loth recessive characters. Similarly, where 

 3 pairs of characters are concerned the F.. ratio will be 27:9:9:9: 

 3:3:3:1. and so on. 



Now so long as the characters of each allelomorphic pair are 

 not only independent in their transmission but are also without per- 

 ceptible effect on each other in the zygotes, such simple ratios will 

 be evident. In such cases types distinct from the parental types 

 will appear, but they will be recognisably due to recombination. 



Difficulty however sometimes arises from the fact that the nature 

 of the allelomorphism is not at first obvious, and a novelty may thus 

 appear which is not at first sight due to mere recombination. 



