PAIRING OF CHROMOSOMES 137 



It is also highly probable, though by no means certain, that the 

 chromosomes which are derived from the male parent remain 

 throughout life distinct from those which are contributed by the 

 female parent. According to this view every ordinary somatic 

 cell has two sets of chromosomes, paternal and maternal respec- 

 tively, and this again is strongly supported by the observations 

 on the germ cells of insects above referred to, where all the 

 chromosomes appear to be duplicated, with the exception of the 

 accessory chromosome in the male animal. 



There is reason for believing, therefore, that in ordinary cases 

 every paternal chromosome in an unreduced nucleus has an equiva- 

 lent or " homologous " mate derived from the female parent, and 

 that the phenomenon of synapsis 1 represents a pairing of these 

 homologous paternal and maternal mates. The reducing division 

 which follows on synapsis consists in the separation of the mates 

 once more, one of each pair going to each daughter cell, so that 

 the matured germ cells are left with a single instead of a double 

 set of chromosomes. 



If we assume that, as seems highly probable, the chromo- 

 somes of each paternal or maternal set are not all identical 

 but differentiated amongst themselves a differentiation which 

 in some cases is actually visible, as shown in Fig. 66 and 

 that one of each kind is necessary to make up the full com- 

 plement of the nucleus of the gamete, the importance of 

 the pairing of homologous chromosomes which takes place in 

 synapsis becomes at once evident, for one of each pair goes to 

 each daughter nucleus, which will therefore be certain to receive 

 a chromosome of each kind instead of a chance assemblage. 

 The chromosome of each kind which it receives, however, may be 

 either the paternal or the maternal representative of that kind, 

 and as these, though essentially homologous, may differ from one 

 another to some extent in accordance with individual peculiarities 

 of the parents from which they were derived, it will be seen that 

 the matured gametes may differ widely amongst themselves in 

 their nuclear constitution. This, as we shall see presently, is a 

 very important matter from the point of view of the theories 

 of heredity and variation. 



This somewhat complex subject will be rendered more readily 



1 The paii-int: of the chromosomes, for which we have used the term "synapsis," 

 is spoken of by some writers as " syndesis," and by others as "conjugation." The 

 use of the latter term seems likely to lead to confusion. 



