CONTROL OF CROSSING-OVER 289 



to be genetically determined. Different forms of Tulipa australis 

 show a different degree of concentration in chiasma frequency, i.e., 

 different interference (v. infra). This modification of chiasma- 

 frequency distribution is probably genetically determined (D. and 

 Janaki-Ammal, 1932). Chiasma formation may also be genetically 

 suppressed (Ch. X). 



Similarly in crosses between Allium species, localisation of 

 chiasmata has to be inherited as a factor difference showing segre- 

 gation (Emsweller and Jones, 1935), and in a normal species. Allium 

 nutans, localisation has appeared in inbred seedlings. Races of 

 Allium fistulosum (Levan, 1933 a) and species of Fritillaria (D., 

 1936) differ in the degree of their localisation, presumably by 

 genotypic control of this property. 



(h) Observations of Crossing-over. The frequency of 

 crossing-over is modified by changes : {a) in external conditions, 

 e.g., temperature (Plough, 1917 ; Kirssanow, 1931, on Drosophila ; 

 cf. Stern, 1928) and irradiation (Muller, 1925) ; (b) in development, 

 e.g., crossing-over varies with the age of the female in Drosophila 

 (Bridges, 1929) ; (c) in genotype, e.g., factors occur reducing 

 crossing-over and modifying its distribution (Detlefsen andClemente, 

 1923 ; Bridges, 1929), or altogether suppressing it (M. S. and J. W. 

 Gowen, 1922, 1928). 



In testing the effect of differences of temperature on chiasma 

 frequency a wider range can be used than in crossing-over experi- 

 ments. Temperatures that are fatal to reproduction are not fatal 

 to the cells in meiosis. Moreover the effect on all chromosomes can 

 be studied at once. White (1934) has been able to show that the 

 variation in chiasma frequencies agrees with that in crossing-over 

 frequencies, but that this variation is shown by the longer chromo- 

 somes with the higher chiasma frequencies only. Whether this 

 effect is a direct one or depends on variation in an interruption of 

 pachytene pairing (as in Fritillaria) we do not know. 



Most changes in crossing-over frequency are differential with 

 regard to different parts of the chromosome, or, we may say, change 

 in frequency is correlated with change in distribution. Thus 

 reduction in crossing-over frequency is attended by an increase in 

 the neighbourhood of the centromere. 



K. A. CYTOLOGY. 10 



