1907.] Structural Constituents of the Nucleus, etc. 457 



real importance of the chromosomes lies more in their structure as bundles of 

 chromomeres than in their material composition alone. 



I have referred to the peculiar circumstance that similar chromosomes pair 

 to form the pseudo-chromosomes. It is, of course, evident that one of each 

 of them will pass into the daughter-cells at meiosis, and hence ultimately into 

 the nuclei of the gametes, which are the direct descendants of these cells. 

 But it is also evident that when the gametes unite at fertilisation, not only 

 will the chromosomes be doubled in number, but that the peculiar ones when 

 present at all will again occur in pairs. The special interest of this lies in the 

 indication that the chromosomes which pair at meiosis are derived from the 

 male and female parents respectively.* This being so, meiosis is seen to be 

 invested with profound significance in relation to the sorting out of the 

 structural elements originally contributed by the parents, and the distribution 

 of entire sets of them, but in different combinations, into the sexual cells. 

 Thus the facts of meiosis are seen to fall completely into line with the 

 conclusions drawn from experiments on breeding as far as the numerical 

 distribution of characters is concerned. 



That this pairing of the chromosomes is a real fact, and not a mere fanciful 

 interpretation put on a complex series of phenomena, is abundantly proved 

 by the work of the last few years. I will give two examples in illustration. 

 It is well known that hybrids are frequently sterile, and this is often found to 

 be associated with the failure on the part of the organism to accomplish the 

 reduction of the chromosomes satisfactorily at the meiotic phase. This may 

 happen if the chromosomes contributed by the two parents are unequal in 

 number, and Eosenberg showed that it occurs in the hybrid between two of 

 our common Sundews. Drosera rotundifolia has 10, and D. longifolia has 

 20 chromosomes in the sexual nuclei. When these two species are crossed, 

 the somatic cells of the hybrid will, of course, exhibit 30 (10 + 20) chromo- 

 somes in their nuclei. Hence, at meiosis, it might be expected that there 

 would be 15 pseudo-chromosomes produced. Instead of this, the number is 

 seen to be 20. But, on examination, the 20 prove to be made up of 10 large 

 and 10 small ones. What has happened is that the 20 contributed by the one 

 parent, as well as the 10 furnished by the other, have retained their autonomy, 

 but the 10 from D. rotundifolia, have united with 10 of the 20 belonging to 

 D. longifolia. The small ones represent the remaining 10, for which there 

 were no mates. The statement that at meiosis the pseudo-chromosomes are 



* I have purposely avoided discussing in detail the peculiar heterochromosomes that 

 have been observed by Wilson and others in insects, as we are not at present in possession 

 of sufficient evidence to fully estimate their possible significance. 



