364 W. J. C. Lawrence Cytoit^ia 2, 



interstitial chiasmata will take even longer to separate, and although 

 these structural differences cannot as a rule be observed in small 

 chromosomes, yet their results may be conspicuous at anaphase. Seen 

 in polar view at anaphase the bivalents which are disjoining first 

 are smaller in size than those which disjoin later, owing to the at- 

 tenuation of the chromatin as the chromosomes are pulled apart. 

 Thus size differences of a critical nature will be apparent at early 

 anaphase, but not at metaphase unless the size distinction is a 

 permanent one. The position of the attachment will also modify the 

 configuration of small chromosomes in side view (cf. Lawrence 1931, 

 p. 292). 



I mention these points because in the absence of structural detail 

 in small chromosomes such differences of configuration and behaviour 

 may assume critical proportions under careful observation. Mid-dia- 

 kinesis is specially suitable for the study of chiasmata (i.e. after termi- 

 nalisation) as these may be clearly seen at this stage, although only 

 with difficulty later. 



Evidence from the Literature 

 a) Cases where secondary association has been observed 



The association of more than two chromosomes at meiosis was first 

 clearly observed by Kuwada (1910) in Oryza sativa L. At metaphase 

 of the first division, the 12 pairs of chromosomes are clearly discernible 

 and the two plates figured present no unusual features. At the second 

 metaphase however a varying number of the 12 bivalents associate 

 intimately in pairs. Kuwada figures three second divisions. In the 

 first (Text Fig. A.e. p. 271) four pairs of bivalents are shown ; in the 

 second (Fig. A.f.) there are probably five pairs and 4 pairs in the re- 

 spective plates. The third figure (anaphase II) shows two closely asso- 

 ciated pairs in each daughter complex. This association is not seen at 

 diakinesis (Plate 8, Figs. 11, 12 and 13). Two somatic metaphases are 

 also shown and in each there is marked association of morphologically 

 similar pairs of chromosomes. This association is often end to end, 

 the pairs appearing in the form of V's or as single chromosomes with 

 median constrictions. 



In 1911 ISHIKAWA reported a similar "pairing" of the chromosomes 

 at the second, but not at the first, metaphase of D. variabilis (2n=64). 

 Inspection of his drawings, however, shows several associations of two 



