METHODS OF ANALYSIS 



299 



2 A. Certain lethal effects, in- 

 cluding segregation in many 

 hybrids, therefore hinder or pre- 

 vent study. 



3. Crossing-over data usually 

 refer to one of the four chroma- 

 tids ; in triploid and attached-X 

 Drosophila, to two chromatids ; 

 in lower plants, especially in 

 fungi, to all four. Except for the 

 last type, therefore, no evidence 

 is obtainable of the relationship of 

 successive crossings-over in the 

 chromosome. 



3A. Interference determinable 

 between crossings-over, and not 

 between chiasmata except when 

 related with pre- and post-reduc- 

 tion observations {Bombyx sex 

 chromosomes; Mather, 1935). 



4. Gene-positions of cross-overs 

 determined from a limited number 

 of genes serving as markers ; 

 certain double and terminal cross- 

 overs are therefore missed (as in 

 male Drosophila) and positions are 

 always approximate. 



5. Crossing-over determinable 

 only in organisms heterozygous 

 for a limited number of genes, and 

 these must be situated in one 

 chromosome. Possible therefore 

 only in the analysis of selected 

 and inbred populations, although 

 equally in male and female cells. 



2A. Effects of X-rays and 

 extreme temperatures and 

 hybridity, giving sterility, can 

 be studied (D., 1932 h ; White, 

 1934 ; Mather, 1934)- 



3. Crossing-over data refer to 

 all four chromatids together. 

 Which pair of the four chromatids 

 have crossed over at a chiasma 

 can be known only in specially 

 favourable material and in certain 

 structural hybrids (D. and Mather, 

 1932 ; D., 1936 d ; Richardson, 

 1936)- 



3A. Interference only determin- 

 able between chiasmata, and not 

 between crossings-over, except 

 when all four ends of pairing 

 chromosomes differ, as in male 

 Drosophila sex chromosomes and 

 dyscentric hybrids (Haldane, 

 1931 ; D., 1934 a, 1936^). 



4. Total number of cross-overs 

 exactly ascertainable not only for 

 given bivalents, but for the whole 

 nucleus (D., 1933 ; Mather and 

 Lamm, 1935). Movement of 

 chiasmata may occur, so that 

 positions are always approximate. 



5. Crossing-over determinable 

 in an unlimited number of species 

 and hybrids, although only in 

 male cells in practice. Equally 

 possible (i) with the extremes of 

 homozygosity and heterozygosity, 

 (ii) with unselected natural popu- 

 lations, and (iii) irrespective of a 

 lethal effect from an abnormal 

 consequence. 



(viii) Conclusion. The direct and parallel evidence now enables 

 us to say that there is a one-to-one correspondence in all organisms 

 studied between chiasmata and crossing-over. Further, we can 

 say that chiasmata behave in all cases as though they were deter- 

 mined by crossing-over between two chromatids of the partner 



