298 CHROMOSOMES IN HEREDITY: MECHANICAL 



toned down. This conclusion, on general grounds, is supported 

 by V. Wettstein's observations (1924) of heterozygosity in diploid 

 spores of Fiinaria when the second division was suppressed. 



The consequences of segregation without reduction have now 

 been shown in two ways. First, Kostoff (1934) has examined the 

 progenies of several hybrids in Nicotiana whose unreduced gametes 

 have functioned in crosses with a third species known itself to be 

 homozygous. These progenies varied, and their variation must be 

 due to crossing-over between unlike chromosomes in the hybrid 

 parent. The same result is found in the tetraploid Fg from a 

 Saxifraga hybrid (Table 26B). Secondly, " mutation " in partheno- 

 genetic plants whose unreduced egg cells develop directly may be 

 ascribed to the same cause (Ch. XI). 



Table 44 



Comparative scope and validity of experimental and direct methods of 

 studying crossing-over. (From D. i935-) 



From Proportions of Combinations 

 in Progeny 



1, Event inferred from its 

 genetical consequences on the 

 assumption, justified by particular 

 and statistical evidence, that the 

 genes are located in structures 

 having the observed properties of 

 chromosomes. Fewer assumptions 

 are made therefore in relating the 

 genetical observations, such as 

 frequency of crossing-over between 

 two genes with the genetical con- 

 sequences. 



2. Crossing-over data usually 

 obtained after a lapse of one 

 generation. There is an elimina- 

 tion of recessive lethal combina- 

 tions in the interval, including 

 those determining absence of 

 crossing-over itself, since where 

 crossing-over occurs it is a con- 

 dition of segregation and fertility. 

 Haploid generation can be studied 

 only in lower plants and Hymeno- 

 ptera. 



From Chromatid Combinations in 

 Germ Mother-cells 



I. Event inferred from conse- 

 quences on the assumption justi- 

 fied by particular and statistical 

 evidence that chiasmata are deter- 

 mined by crossing-over. Fewer 

 assumptions are made therefore 

 in relating the cytological observa- 

 tions, such as frequency per 

 chromosome, with the mechanical 

 conditions (D., 1934 ^)- 



2. Crossing-over data obtained 

 after a lapse of a few minutes, 

 during which no elimination can 

 occur. 



