496 CELL MECHANICS 



found instead of terminal associations in organisms with nonnally 

 complete terminalisation (CEnothera, Primula, Campanula). 



6. The frequency of terminal metaphase associations of fragments 

 is in proportion to the frequency of interstitial chiasmata found in 

 large chromosomes, where the fragments therefore must usually 

 develop one chiasma [Fritillaria imperialis). 



7. Further, these fragments are associated at only one end, as 

 they should be, having only one chiasma formed, although the 

 larger chromosomes are associated at both ends, having several 

 chiasmata formed (Tradescantia, Datura, Blakeslee, 1931). 



8. The terminal association of tetraploids and triploids does not 

 consist in union of all homologous ends, but in a random assortment 

 of possible associations of such homologous ends in every possible 

 combination such as would be determined by chiasmata from 

 randomly paired threads, as observed in a polyploid at zygotene 

 (Hyacinthus and Tulipa), diplotene [Primula, Campanula) and 

 metaphase [Datura). 



9. Evidence from polyploids which agrees with the assumption that 

 all chromosome association at meiosis is by chiasmata, necessarily 

 implies that a terminal association is a chiasma. This evidence is 

 the degree of, and variation in, the frequency of multivalent 

 association of chromosomes (sometimes of different lengths) in 

 triploids and tetraploids already described [Hyacinthus, Lilium 

 tigrinum, Primula sinensis). 



10. Similarly, the association of chromosomes in hybrids whether 

 terminal or interstitial agrees in its frequency and in its variation 

 between different nuclei of the same individual and between the 

 nuclei of different individuals, as is expected from the conditions, 

 frequency and variation of chiasma formation [Triticum, Crepis, 

 etc., cf. Mather, 1935 b). 



These considerations, as we have seen, are of primary importance 

 in interpreting chromosome behaviour at meiosis. Chiasmata are a 

 condition of metaphase pairing, and the understanding of their 

 structural properties is a condition of understanding how this 

 pairing comes about. They are also, as we have seen, of importance 

 in interpreting meiosis itself. They show that specific attraction 

 operates only between pairs of particles at all stages of meiosis. 



