134 



ELIZABETH W. DAVIES 



as yet never been found growing with any of the other species in this country. Hence 

 no hybrids have been recorded. 



(6) C. hostiana and C. distans 



This study would not be complete without mentioning- C. hestiana DC. and C. 

 distans L., two members of the closely related section Distantes, for the former species 

 very frequently hybridises with C. lepidocarpa and C. demissa. C. hostiana (n = 28), 

 like C. demissa, grows in a wide range of habitats and tolerates many different soil con- 

 ditions, which probably accounts for the frequency of its hybrids. 



The hybrid between C. hostiana and C. lepidocarpa is sufficiently well known to 

 have a name (C. X xanthocarpa), and is of fairly frequent occurrence. It grows (Table 2) 

 at Chippenham Fen, Cambridgeshire, where the h^^brid forms a large clone, which appears 

 to multiply entirely vegetatively and now covers a considerable area, surrounded by the 

 parent species. 



Likewise C. hostiarm hybridises with C. demissa. This hybrid was collected by Loch 

 Tummel, Perthshire, and the plant showed a highly irregular meiosis with several chains 

 of chromosomes and numerous univalents. 



Lastly a hybrid between C. distans (n = 37) (Plate 13, fig. 15) and C. lepidocarpa 

 was recorded from Cow Common, Totternhoe, Bedfordshire. This hybrid formed a 

 clonal stand hke C. X xanthocarpa, and had highly sterile pollen and a very irregular 

 meiosis. 



Thus it seems that all the members of this aggregate are capable of hybridisation, 

 and considerable gene-flow between species is possible. It seems also that it is not 

 genetical incompatibility but ecological and topographic barriers that keep C. flava, 

 C. lepidocarpa, C. demissa and C. scandinavica apart, as distinct entities, while, in contrast, 

 late seasonal periodicity seems to play the major role in the case of C. serotina. 



(b) Artificial Hybrids 



These flavoid species of Carex, like all those in the subgenus Carex, lend themselves 

 to artificial hybridisation, as the male spike is easily removed without damaging the 

 female organs. The crosses were carried out during }vlay and early June, and the ripe 

 seeds collected at the end of July and sown immediately. 



The artificial hybridisations included all possible combinations between the five 

 British members of the aggregate, and some crosses between these and C. mairii, C. 

 extensa and C. hostiana (Table 3). However, although in most combinations some 

 apparently ripe seed was produced, germination is slow and spasmodic in this genus, so 

 as yet only a few of the hybrid seedlings have produced fertile spikes. These are listed 

 in Table 3, and Plate 13, fig. 20 shows the very irregular meiosis of these artificial hybrids. 



Thus it would seem that these five British species included in the C. flava aggregate, 

 are capable of considerable gene-flow from one to another, and produce plants with 

 between 20 and 30 per cent good pollen. They should therefore be regarded as ecospecies, 

 and the group as a whole as one coenospecies (Gregor, 1939; Clausen, Keck & Hiesey, 

 1939). 



6. Discussion 



The cytology, evolution and origin of the aneuploid series in the genus Carex will 

 be discussed more fully in another paper. 



However, it seems reasonable to postulate from the present c\tological, ecological 

 (Davies, 1954) and taxonomic (Davies, 1953a, b, c) evidence that, within this coenospecies. 



