366 W. J. C. Lawrence Cytoiogia z, 



In the meantime Meurman (1929) LAWRENCE (1929) and Darling- 

 ton (1929, 1930) had abandoned the view that secondary association was 

 a distinct phenomenon, attributing the apparent absence of material 

 connections between associated bivalents to the extreme fineness and 

 attenuation of the threads in these small-chromosomed plants. 



Further study of meiosis in Dahlia however has provided evidence 

 which, in conjunction with other workers' results, establishes the 

 reality of this second type of association of homologous chromosomes.' 

 A brief discussion of secondary association was given in my papers on 

 Incompatibility (1931) and Dahlia (1931). Further evidence as to its 

 reality is supplied by the work of my colleague Mr. A. A. Moffbtt on 

 the Pomoideae (1931). Thirty-one species from eleven genera all show- 

 ed pronounced secondary association of the bivalents. Twenty-two 

 species were "diploids" (n=17) and Moffett notes that "in the great 

 majority of divisions examined at diakinesis seventeen bivalents were 

 observed." The number of groups observed at metaphase is never less 

 than seven with never more than three bivalents per group. Maximum 

 association is three groups of 6, and 4 groups of 4 chromosomes. The 

 cytological evidence from Pyrus and other forms suggests that the "dip- 

 loid" species are secondary polyploids with 7 as the basic chromosome 

 number, three chromosome types being represented 6 times and four 

 types four times. The frequency of the respective groups at meta- 

 phase leads Moffett to the conclusion that secondary pairing in the 

 Pomoideae is determined by the constant qualities of the chromosomes 

 and is not merely an accident of fixation. It is highly significant that 

 Darlington and Moffett (1930) observed primary association to be 

 more frequent in " triploid " than in " diploid " apples. Thus chromo- 

 somes whose homology could only be inferred in the diploid by their 

 secondary association, are shown under suitable conditions (i. e. in the 

 absence of competition in the triploid) to be identical in part, by their 

 primary association. The evidence from Pyrus therefore is proof of 

 the homology of chromosomes which associate secondarily. 



b) Cases in which secondary association may be inferred 



Evidence of secondary association in the literature is so profuse 

 that reference can only be made to the more obvious examples. The 

 value of this evidence from the drawings of other workers is, of course, 

 mainly dependent on an unknown factor— the quality of fixation. This 



