76 



ELIZABETH W. DAVIES 



means of distinguishing C. serotina and the rather similar C. scandinavica, for although 

 in contrast to C. flava the leaves of both these species are narrow and canaliculate, those 

 of C. scandinavica are always shorter than the stems. 



C. demissa has broad leaves, about three-quarters the length of the culms, while 

 the leaves of both subspecies of C. lepidocarpa are short and comparatively narrow. Subsp. 

 lepidocarpa has leaves that are about half as long as the stems, while those of the subsp. 

 scotica E. W. Davies are even shorter. 



(b) Width 



The leaves of C. flava and C. demissa in this order are broadest, and usually flat, 

 while both subspecies of C. lepidocarpa have narrower leaves. However, those of subsp. 

 lepidocarpa are grass-like and herbaceous, while subsp. scotica has tough fibrous leaves, 

 which retain this texture in cultivation, a character which is associated with the severe 

 climatic conditions in the north and at high altitudes. The leaves of C. serotina and 

 C. scandinavica are very narrow, and in contrast to all other members of this group they 

 are inrolled. 



3. Spikes 



(a) and (b) Arrangement and number of male and female spikes 



The arrangement and number of the spikes are reliable characters, which show 

 little phenotypic variation under normal conditions, and remain constant in cultiva- 

 tion. However, occasionally irregularities have been noticed when culms are produced 

 as a second crop at the end of the season. This is unusual and these late fertile shoots 

 are unreliable for identification. 



As can be seen from Table 1, and in Plate 8, C. flava and C. serotina show a similarity 

 in the number and arrangement of their spikes. Both species have a comparatively short 

 sessile male spike, which is occasionally partly female, while the female spikes (3-4 in 

 C. flava and 3-5 in C. serotina) are usually contiguous and crowded round the male 

 spike. In contrast the male spikes of C. lepidocarpa, C. demissa and C. scandinavica are 

 considerably longer and the female spikes (1-3 in C. lepidocarpa, 2-4 in C. demissa and 

 2-3 in C. scandinavica) are often separated from the male spike, and sometimes from 

 each other (Fig. 2). In C. demissa there is frequently one remote basal spike with a 

 long peduncle (Fig. 2d). 



4. Perigynia or Utricles 



(a), (b) and (c) Length and shape of perigynia and length of beak 



The dimensions and shape of the perigynia have provided the most reliable biometric 

 character, as they show very little variation and are seldom influenced by environmental 

 conditions (Plate 9). In fact it is possible to identify the members of this group from 

 the length of the perigynia alone. For ahhough a slight overlap can be seen in the 

 accompanying histograms (Fig. 1), in most well-grown specimens there is at least a 

 difference of 1 mm. between any two of the species. Thus the perigynia lengths are 

 as follows :- 



C. flava 



C. lepidocarpa ssp. lepidocarpa 



C. lepidocarpa ssp. scotica 



C. demissa 



C. serotina 



C. scandinavica 



mode 6-0, range 5 -25 -7-0 mm« 

 mode 4-25, range 3-5 -4-5 mm- 

 mode 4-5, range 4-0 -5-0 mm. 

 mode 3-25, range 3-0 -3-5 mm. 

 mode 2-25, range 2-0 -2-5 mm. 

 mode 1-75, range 1-0 -2-25 mm. 



