154 STRUCTURAL HYBRIDS 



an ancestor. There is no direct evidence to show whether the 

 interchange gave rise to the heterozygote at once in nature or 

 whether it gave rise to a new homozygous race which afterwards by 

 crossing gave rise to the hybrid. The second is obviously the 

 source of those ring-forms that have appeared in experiments. 

 There is some reason to suppose, on the other hand, that where they 

 occur in nature their preservation is due to the eUmination of the 

 homozygous products of the original interchange hybrid (v, 

 CEnothera), i.e., that they arose directly from the change without 

 crossing between different parental zygotes. 



In experiments with Datura, Pisum and CEnothera, hybrids with a 



Fig. 53. — A ring of twelve arranged disjunctionally at first meta- 

 phase in Rhoeo discolor. X 2800. (From D., 1929 c.) 



ring of four have been obtained by crossing different homozygous 

 races. The parents evidently differed in respect of one interchange, 

 and the offspring (of the type AB-BC-CD-DA) may therefore be 

 described as " single-interchange " heterozygotes. Hybrids with a 

 ring of six or two rings of four are similarly " double-interchange " 

 heterozygotes, and they have been produced either by crossing two 

 parents which differed in two interchanges [AB, CD, EF and 

 BC, DE, FA) as in Datura, or by crossing two single-interchange 

 heterozygotes, which have one gametic type in common (as in 

 Pisum, Zea and Campanula). From such a cross four kinds of 

 progeny will result. Thus (AB-BC-CD-DA, EF-EF) x (AB-AB, 

 CD-DE-EF-FC) will give homozygotes with simple pairing (i.e., 

 AB-AB, CD-CD, EF-EF), two kinds of " single " heterozygote 



