II] MATERIAL BASIS 149 



made to meet this difficulty, of which perhaps the 

 most adequate is that the chromosomes are not in- 

 divisible entities, but are composed of smaller units, 

 each of which corresponds with one Mendelian factor. 

 The chromosomes are not permanently present in 

 the distinct form which is seen during cell-division, 

 but during the resting condition of the nucleus their 

 substance becomes diffused over a network of threads, 

 only to be collected again into definite chromosomes, 

 having the same number and form as before, pre- 

 paratory to the next division. If each chromosome 

 consists of a series of units having a definite arrange- 

 ment, and these units become scattered in the 

 'resting phase,' but are re-collected in the same 

 order when the chromosomes are re-formed, it does 

 not seem unlikely that a unit N may take the place 

 of the corresponding unit n from the other chromo- 

 some of the pair, so that if the chromosome A 

 consisted at one division of units M, N, O..., and the 

 corresponding chromosome a consisted of m, n, o..., 

 after the resting stage N and n might have exchanged 

 places, and chromosome A would consist of M, n, O. . 

 and a of m, N, o.... By some process of this kind 

 it seems probable that the observed phenomena of 

 chromosome reduction would account for all the 

 facts of Mendelian segregation. 



It must be stated quite clearly, however, that 

 the study of the possible relation between chromo- 



