252 INTRODUCTION TO CYTOLOGY 



folia (20 chromosomes) ' is crossed with D. longifolia (40 chromosomes) 

 there results a hybrid with 30 chromosomes, of which 10 are contributed 

 by rotundifolia and 20 by longifolia. When synapsis occurs preparatory 

 to reduction in this hybrid only 10 bivalents are formed, 10 chromosomes 

 remaining unpaired. This was taken by Rosenberg to mean that the 10 

 rotundifolia chromosomes pair with 10 of the longifolia ones, leaving the 

 other 10 of longifolia without synaptic mates. Had any chromosome of 

 the duplex group of 30 been free to pair with any other, 15 bivalents 

 would have been produced. 



Other instances of this phenomenon may be mentioned. By crossing 

 (Enothera Lamarckiana (seven chromosomes in gamete) with (E. gigas (14 

 in gamete) individuals with 21 chromosomes are obtained. Geerts 

 (1911) found that, preparatory to reduction, the seven Lamarckiana chro- 

 mosomes pair with seven of the gigas chromosomes, leaving the other 

 seven of gigas unpaired. On the contrary, however, Gates (1909) found 

 that the 21 chromosomes in a lata-gigas hybrid simply separate into two 

 approximately equal groups, usually of 10 and 11 chromosomes re- 

 spectively. Kihara (1919) reports that in some 35-chromosome wheat 

 hybrids formed by crossing Triticum polonicum (14 chromosomes in 

 gamete) with T. spelta (21 in gamete) there are present in the heterotypic 

 prophase 14 bivalents (polonicum conjugated with spelta) and seven 

 univalents (spelta). The 14 bivalents are arranged on the spindle and 

 separate as usual, whereas the seven unpaired spelta chromosomes split 

 longitudinally at the first mitosis and distribute themselves irregularly 

 at the second (Fig. 100). An analogous condition is found in Pigcera 

 hybrids by Federley (1913). 



A very significant additional suggestion with respect to synapsis was 

 made by McClung (1900) and Sutton (1902): not only are the two chromo- 

 somes which conjugate derived from the two parents, but they are hom- 

 ologous each chromosome of one parental set pairs with a particular 

 chromosome of the other parental set, the two members of the resulting 

 bivalent being presumably of corresponding hereditary value, as will be 

 shown in Chapter XV. The evidence for this important hypothesis was 

 found chiefly in Brachystola (Fig. 101) and a number of other insects 

 having chromosome complements made up of members with constant 

 characteristic differences in size and shape. Many such cases have been 

 subsequently discovered, especially by McClung and his eoworkers in 

 their extensive researches on insect spermatocytes. As examples among 

 plants may be cited Crepis virens, Najas major, N. marina, and Vicia 

 faba. 



Crepis virens (Rosenberg 1909) (Fig. 102) has six chromosomes: two 

 long, two medium sized, and two short. When synapsis occurs the like 

 chromosomes pair, forming bivalents of three sizes. The members of 

 each pair separate and pass to the daughter cells at the first maturation 



