HYBRIDISATION OF EUPHRASIA 



267 



number. He envisages repeated back-crossing leading to fertility, resulting in intro- 

 gression by which one or other of the original species may be progressively modified. It 

 would seem, however, that introgression into the diploid, by this process, is of rare 

 occurrence. 



No definite indications of introgression from diploid into tetraploid have yet been 

 encountered in Euphrasia. / 



The conclusions drawn with regard to hybridisation mean that gene exchange is 

 possible between two groups which would be expected to be genetically isolated on 

 account of polyploidy. In Turesson's terminology of experimental taxonomy, forms 

 that would be expected to belong to two coenospecies belong to the same one. A difference 

 of chromosome number may in some cases mark the separation of two coenospecies, while 

 in others, such as that of Euphrasia, it will not. It is therefore perhaps preferable to use 

 the new terminology of Gilmour & Heslop Harrison (1954), which in its general definitions 

 is explicitly elastic, but which is made precise in any particular case by the qualifications 

 introduced by the user. According to this terminology E. anglica and E. micrantha may 

 be said to constitute two different hologamodemes (within which all individuals are able 

 to interbreed with a high level of freedom), which are not completely discontinuous and 

 therefore belong to one coenogamodeme. Gene exchange between the two hologamodemes 

 is at present only known to go in one direction. In a particular population E. anglica and 

 E. micrantha form two gamodemes which are largely, but not completely, isolated genetic- 

 ally, while the diploid hybrid forms another gamodeme genetically isolated from E. 

 micrantha (which may be called a cytogamodeme) and ecologically isolated from E. 

 anglica (which may be called an ecogamodeme). Though the frequency of transfer of 

 genes from the tetraploid to the diploid population is low, individuals showing inter- 

 mediate characters may be very numerous. In spite of the abundance of intermediate 

 individuals and the possibility of their closely resembling the tetraploid parent, the 

 two gamodemes will be able to persist as largely discontinuous breeding units because 

 of the difference in their chromosome numbers, which cannot be bridged by intermediates 

 as the morphological differences can. 



Finally, attention is drawn to the occurrence of hybrid forms with flowers larger 

 than those of either parent. This is so frequent as to suggest that such forms may be 

 selected for some reason. 



Summary 



The parents of some reported hybrids are now known to be diploid and tetraploid 

 respectively. The hybrids, however, are fertile and numerous. 



Such a hybrid is described from Withypool, Exmoor. It looked intermediate between 

 E. anglica and E. micrantha which accompanied it, and it grew in longer grass than the 

 dwarf diploid E. anglica. It was diploid. A single triploid individual was also found, 

 evidently a hybrid between E. anglica and E. micrantha. 



Probably the diploid hybrid arose through the back-crossing of a triploid to the 

 parental diploid by means of a haploid gamete. If this hybrid possesses characters 

 derived from the tetraploid which enable it to grow in a habitat not occupied by the 

 parental diploid, it may form an independent population of a new diploid type with 

 distinct morphology and habitat. 



The plant described as E. vigursii, and regarded as E. anglica X micrayitha by Pugsley, 

 was investigated, and presented an analogous situation, but no triploid was found. E. 

 vigursii, though morphologically different from the Withypool plant, appears to have 

 the parentage suggested by Pugsley. E. vigursii occurs in three vice-counties, is distinct 

 in morphology and habitat, and resembles a species. 



