488 
L. DONCASTER. 
understanding of the way in which the facts referred to are 
related to the general question of the material basis of liere- 
ditary transmission and of sex-determination. 
The argument falls naturally into the two divisions of the 
relation of the chromosomes to Mendelian heredity on the one 
hand and to sex on the other, but these two are intimately 
connected hy the phenomena of sex-limited inheritance, from 
cases of which some of the most important evidence with 
regard to both branches has been obtained. It will, there- 
fore, be most convenient to consider in their relation to 
chromosome behaviour, first, heredity alone, then sex, and 
finally the phenomena of sex-litnited inheritance. 
(1) Chromosomes and Mendelian Factors. 
The opinion that the chromosomes are the ‘M^earers of 
characters which show Mendelian inheritance has been 
supported on several quite distinct grounds, some of which 
are inferential, others depending on direct observation. The 
earlier arguments in favour of this idea were founded entirely 
on observations which suggested that the behaviour .of the 
chromosomes in the maturation of the gametes is of exactly 
the kind which would give rise to Mendelian segregation. 
The facts are so familiar that no detailed account is needed. 
It has been maintained by many observers, especially in cases 
ill which the chromosomes differ conspicuously among them- 
selves in size, that the nuclei of the somatic cells and sperma- 
togonia or oogonia contain a double complement, composed 
of sets of maternal and paternal origin respectively, and that 
in synapsis” each maternal chromosome first pairs with a 
corresponding' paternal one, and then separates from it at one 
of the maturation divisions into a different daughter-cell. If, 
then, each chromosome corresponded with a Mendelian 
character, and if chromosomes bearing allelomorphic 
characters always paired together, the segregation of the 
members of an allelomorphic pair into different gametes 
would be accounted for. And since it is a matter of chance 
with regard to any pair whether the paternal or maternal 
