256 



P. F. YEO 



produced by self- or cross-fertilisaton, will be highly variable. Thus a great source of 

 variation in the diploid becomes available. 



For this process to result in the production of a distinct form such as E.187A, there 

 must be close at hand a plant community not occupied by a diploid species, to which 

 the progeny of the back-crossed individual can gain access.. This community must be 

 one to which these progeny are adapted or can become adapted, by selection from their 

 especially great variation. The genes derived from the tetraploid would thus enable 

 their possessors to evade the parental diploid and grow in ecological isolation, and so 

 obtain sufficient genetical isolation to remain distinct. The habitat of relatively long grass 

 at Withypool fulfils this requirement. 



Some additional possibilities can now be stated. One is that the diploid may result 

 from a triploid fertilising itself by producing both male and female spores that are viable. 

 There arise various other considerations which do not essentially alter the picture. The 

 process of hybridisation between diploid and tetraploid, followed by back-crossing to the 

 diploid, is presumably happening all the time with some low frequency. The process 

 may also occur directly between the tetraploid and the new diploid. By both these 

 means, the new form will be able to draw into itself other genes as they become available, 

 provided they are favoured by selection. One would expect that ultimately, when the 

 new form had " sampled " all the genes the tetraploid had to offer, an equiUbrium would 

 be reached. Selection after the original cross, and after later ones, would reduce variation, 

 but could result in some variation persisting in response to minor habitat variation. 



It is thus possible to explain the occurrence of a fertile diploid whose parents are 

 diploid and tetraploid respectively, and to account for its existence as a morphologically 

 distinct population. The hypothesis would appear still more probable if the suggested 

 sequence of hybridisations could be made to occur experimentally. So far efforts in this 

 direction have not even resulted in the production of a triploid. Finally it may be mentioned 

 that E. con/usa Pugsl. also grew on the heath at Withypool and approached quite near to 

 the colony of the hybrid. Some individuals of the hybrid show suggestions of £. confusa 

 in their branching, and it is conceivable that this species has contributed to its origin 

 by the same process as suggested for E. micrantha. 



Euphrasia vigursii Davey 



This is the commonest and best known of the reported hybrids whose parents are 

 respectively diploid and tetraploid. It was described as a species by F. H. Davey (1907), 

 and was said to occur at Forth Towan and Perranzabuloe, both in v.c. 1, W. Cornwall, 

 and at Roborough Down, v.c. 3, S. Devon. It is now well represented in herbaria, by 

 specimens not only from the original localities, but from several others in Devon and 

 Cornwall. Pugsley (1930) regarded it as a hybrid between E. anglica and E. micrantka. 

 From the morphological point of view, this seems plausible, and Pugsley pointed out 

 that in some gatherings a proportion of the capsules are malformed, though such capsules 

 are not necessarily completely barren. 



Observations on E. vigursii at Roborough Down 



The following observations were made at Roborough Down, which is a tract of 

 heathland extending along, and on both sides of, the Plymouth-Tavistock road, for a 

 distance of about six miles. The area investigated extended for about three-quarters 

 of a mile from the southern end, and was mostly to the west of the main road (grid ref. 

 20/5063-20/5064). 



The Euphrasiae present were E. confusa (of which three very small colonies were 

 found), E. anglica, E. vigursii, and a form which will be referred to as E. nemorosa, though, 



