CJIKOMOSOMES, HEREDITY AND SEX. 
489 
element goes to a particular daughter-cell, the iudepeudence 
of different allelomorphic pairs in the formation of gametes is 
also explained. The mechanism for producing Mendelian 
segregation appears so perfect that it is difficult to believe 
that the two phenomena are unrelated, but there are serious 
difficulties which have first to be explained. The first of 
these is that there may be more pairs of allelomorphic 
characters than there are pairs of chromosomes, and yet the 
characters of different pairs show no constant association with 
one another. Several theoretical suggestions have been 
made to meet this objection, most of Avhich assume that the 
chromosomes are not indivisible units, but are made up of 
smaller parts, each of which is the bearer of one Mendelian 
character. By some it is supposed that the units may be 
interchanged during synapsis ; by others, that they become 
separate in the ^ffi-esting^^ nucleus, and that it is a matter of 
chance whether they return into one or other of the homo- 
logous chromosomes to which tliey belong. The evidence for 
the compound nature of chromosomes is now so strong^ that 
the difficulty cannot be regarded as very serious, but the exact 
manner in which the units are arranged in the chromosomes 
is far from being settled. One hypothesis with regard to 
this part of the question will be discussed more fully later, in 
connection with gametic coupling. 
The second objection to the hypothesis that the pairing and 
separation of chromosomes in gametogenesis gives rise to 
Mendelian segregation is more serious, since it is based on 
the denial that the chromosomes behave as described. Some 
observers refuse to credit the chromosomes with individuality 
of any kind (33), and without some sort of individuality, 
1 Compound chromosomes have been described in Ascaris (T. 
Boveri, ‘ Ergebnisse iiber die Konstitution der chromatischen Snbstanz 
des Zellkerns,’ Jena, 1904, p. 27); in the bee (H. Nachtsheim, 'Arch. 
Zellforsch.,’ xi, 1913, p. 169) and in other cases. The writer finds that 
in the nearly related moths, Nyssia zonaria and Biston hi]-taria, 
the chromosomes of the former are exactly four times as numerous as ' 
in the latter, and much smaller, so suggesting that those of B. 
hirtaria may be compound. 
