SOMATIC PAIRING 235 



to a greater or less extent on the periphery of the plate according 

 as the repulsion of the poles is greater or less ; (iii) the bodies 

 of the chromosomes lie as far apart from one another as their 

 attachment to their centromeres allows. When the bodies of 

 the chromosomes are long, those that are in the middle therefore 

 lie on both sides of the plate and those that are on the edge lie off 

 the plate and can be seen to move in the currents of the cytoplasm. 

 These rules argue an equal repulsion of the two poles for the centro- 

 meres and a mutual repulsion of the bodies of all the chromosomes. 

 The rules apply to meiosis when we make allowance for there being 

 in each bivalent two centromeres which repel one another in the 

 axes of the spindle and therefore like the anaphase daughter centro- 

 meres of mitosis lie on either side of the plate. 



In some organisms the even distribution of the chromosomes on 

 the metaphase plate is modified by another factor. 



At mitosis in the diploid in many Diptera, pairs of chromosomes 

 are seen to lie specially close together, although still never touching, 

 and these are found to be similar pairs, with similar parts lying 

 parallel. This is due to a specially exaggerated property of attrac- 

 tion in this group [cf. Metz, 1916, igzGetal. ; also Fig. 116). The 

 same somatic pairing has been found between the chromosomes of 

 polyploid plants at mitosis (e.g., Dahlia, Lawrence, 1931). In this 

 case more than two chromosomes being attracted to one another, 

 the groups lie radially instead of parallel. 



More striking, however, are the conditions in doubled nuclei 

 which have arisen through the failure of the chromosomes to 

 separate at a preceding division {e.g., Spinacia, Stomps, 1911 ; 

 Sorghum, Huskins and Smith, 1932 ; Apotettix, Robertson, 1930 ; 

 Iberis, Manton, 1935). Probably in these the daughter chromosomes 

 have remained together during the resting stage, and have therefore 

 been in a suitable position to exercise their special attraction on 

 one another during the prophase. At metaphase every pair is 

 distinguishable by the chromosomes being of similar shape and 

 lying parallel. This is evidence not only of the method of origin 

 of polyploid nuclei, but also of the static condition of the chromo- 

 somes during the resting stage, and of their attraction at metaphase. 



The special attraction of chromosomes seems to be specific to 



