Manchester Memoirs, Vol. xlviii. (1904), A^t". ^4- n 



this fact is recognised by naming such characters allelo- 

 morphs. It is hardly necessary to say the germ contains 

 not merely one allelomorph (as we have been imagining 

 for the sake of simplicity) but very many : in fact Mendel 

 recognised 7 such pairs in his peas, and this must be the 

 merest fraction of the actual number that exists. 



{c) We are thus indirectly led to the conception of 

 ''^compound characters, borne by one gamete, transmitted 

 entire as a single character so long as fertilization only 

 occurs between like gametes." (I.e. p. 29.) The reader is 

 here strongly advised to refer to the first thirty-five pages 

 of Bateson's book. (Bateson :02^.) 



(8) The relation between these new conceptions 

 and the theory of the origin of species by 

 discontinuous variation, {a and b ; gametic 

 purity and unit-characters?) 

 Let us suppose a new characterto arise bydiscontinuous 

 variation ; for example the possession of a trunk in a race 

 of previously snoutless elephants. According to the 

 statistical view of heredity this new character would be 

 soon swamped by the, so to speak, normal trunkless 

 ancestry of the new form itself and of its mate ; for one of 

 two courses is open to the new form ; it may either unite 

 with another like it, which would depend on a series of 

 contingencies : — the production of two such beasts at 

 the same time, at the same place, of opposite sexes, 

 and the condition that they were not averse to one 

 another. Or it might unite with a trunkless relative. But 

 even in the former case the offspring would have a smaller 

 trunk than its two parents. And if this smaller trunk 

 were to be perpetuated its owner would have to unite with 

 another trunk-bearing variety which therefore would have 

 to arise at the proper time, be of the opposite sex and in 



