562 THE PURE LINE HYPOTHESIS. 



any modification of the potentialities or factors in the g-erm-cells, 

 and are consequently non-inheritable. If a factor itself is changed 

 by any cause, a variation of another nature results in the character, 

 and such a variation is called a mutation. In actual practice it 

 i.5 impossible to distinguish between a so-called casual or fluctua- 

 ting variation and a mutation except by the test of breeding, since 

 it is becoming increasingly obvious that minute variations may be 

 inheritable. 



2. The expression " character " in this connection has a 

 technical significance, since it generally comprises a group of 

 characters which are inherited as a whole. An inheritable 

 character, or group of characters, represented by a specific 

 potentiality or factor in the germ-cell, is called a " unit character." 



3. In crossing varieties which exhibit some striking difl:'er- 

 ences, it is found by breeding experiments that quite frequently 

 the corresponding characters in the two organisms refuse to 

 mingle intimately in the offspring, and on mating these offspring 

 together, or by self-fertilisation, the characters of the grand- 

 parents reappear in a practically tmaltered condition, and in a 

 definite ratio. 



Cytological research appears to show that the chromosomes 

 of cells are arranged in a definite number of homologous pairs, 

 one of each pair being derived from the e9"g-cell and one from the 

 sperm-cell. It is supposed that an inheritable character of a homo- 

 zygous individual is represented by two similar factors in the cells 

 of the body and in the immature germ-cells, one being lodged in 

 each member of a pair of homologous chromosomes. With this 

 supposition, the seoaration of the chromosomes in the various 

 mitotic divisions of the sex-cells, and their subsequent coupling 

 in fertilisation, explain in an adequate manner the typical Mende- 

 Man 'ratios and other phenomena associated with this type of 

 inheritance, which may be called Non-blending Inheritance (see 

 plate). 



4. Non-blending Inheritance is particularly noticeable in the 

 case of characters distinguishing breeds and varieties which 

 have arisen during domestication ; but even in such characters 

 blending may take place. The variations which occur anions' the 

 individuals of a natural species do not usually exhibit this kind 

 of inheritance. As a rule the characters become intimatelv 

 united on crossing, and no segregation occurs in any subsequent 

 generation: this is Blending Inheritance. It would certainly 

 appear that in Non-blending, or Mendelian, Inheritance, the 

 factors for the characters are contained in the chromosomes, but 

 in the case of characters which do not exhibit the Mendelian 

 relationship there is the possibility that the potentialities or 

 factors are borne by the cytoplasm. However, certain artificial 

 fertilisation experiments among Echinoderms would seem to 

 indicate that the cytoplasm, as a rule, does not bear the factors 

 for hereditary characters; but some contradictory results have 

 been obtained, and further experimental work with Echinoderm 

 eggs is very necessary. 



