PROLIFERATION OF SPIKELETS IN BRITISH GRASSES 



51 



earliest reference found by the present author, in which pseudalpina is given as a variety 

 of D. cespitosa, and not of A. cespitosa. Thus, strictly, D. cespitosa (L.) Beauv. var. 

 pseudalpina (Syme) Druce is correct. 



Seminiferous specimens of D. alpina are rare, although Linnaeus never mentioned 

 vivipary in A. alpina, Don sent a flower-bearing plant to Smith (1811). Parnell (1842) 

 describes and illustrates only seminiferous forms. Nygren (1949) received a sexual seed- 

 bearing plant. The present author found a plant most of whose spikelets bore florets 

 only. Another plant in culture (see below) produced a completely flower-bearing culm, 

 although strongly viviparous when gathered in Harris. 



7.2. Morphological observations 



During the course of this study no viviparous specimens have been seen with a 

 basally inserted awn. The awn is sometimes inserted about the middle of the third 

 glume (i.e. the lowest lemma); this is the typical D. alpina arrangement. Often the 



Fig. 7. Deschampsia spikelets. 

 A. Deschampsia caespitosa seminiferous (x5'5): 1. lower floret; 2. upper florets. B. Deschampsia 

 cespitosa var. pseudalpina ; C. Deschampsia ' intermediate ' viviparous type. D. Deschampsia alpina vivipar- 

 ous ; 1. lower glumes removed to display the poorly developed floret (B, C and D are all X 1-4). 

 E. Deschampsia alpina seminiferous (x 7) ; 1. lower glumes removed. 



third glume is so metamorphosed that the awn appears as a terminal bristle or is absent. 

 This, being the only other widespread form, is presumably the D. cespitosa var. pseudalpina 

 condition. 



There is an intermediate group, in which the awn is a hair lying between the twin 

 ear-like terminal lobes of the third glume. In these intermediate forms there is a very 

 short subterminal awn which is not clearly separated from the glume ; it neither appears 

 ' to arise ' nor ' to be inserted.' Argument for inclusion with either of the main forms 

 may be advanced, perhaps more strongly for D. alpina. In Figure 7 these types are 

 illustrated. 



The specimens in the herbaria of Linnaeus and Smith (Linn. Soc. Lond.) and of Syme 

 (B.M.) were consulted. There are no specimens labelled A. alpina by Linnaeus but in 

 his collection are two viviparous plants, apparently D. alpina with slight intermediate ten- 

 dencies, labelled * Lapp ' and joined to A. cespitosa. 



The plants in Smith's collection appear to be a seminiferous D. alpina labelled A. 

 alpina; viviparous 'intermediate' plants marked 'A. laevigata,' which he later con- 

 sidered A. alpina; and finally a viviparous plant of var. pseudalpina placed under A. 

 I cespitosa. (To avoid damage the specimens could not be very closely examined.) 



