BIOSYSTEMATIG STUDY OF SOME GLYCERIA SPECIES -I 



295 



From these data it is clear that the G. plicata grains are, on average, larger than those 

 of the other two species. A " t " test shows that this difference is significant at P = 0-05. 

 The frequency distribution of pollen grain size is shown in Fig. 2, where histograms of 

 the three species are superimposed, those for G. declinata and G. fluitans*hemg more 

 or less coincident. In each case there is a fairly good approximation to a normal curve. 

 Because of the overlapping ranges of variation, a statistical test would be required in 

 order to separate G. plicata by means of pollen size. Fitzpatrick (1946), whose results 

 are given above for comparison, found that both the tetraploid species had significantly 

 larger pollen grains than those of the diploid, but only 20 of the largest and fullest grains 

 were measured, which suggests that this result may not be reliable. 



Pollen fertility, determined whilst making the foregoing measurements, was good 

 in all three species. 



The hybrid G. X pedicellata 



G. fluitans and G. plicata can easily be distinguished morphologically, but their 

 hybrid shows a wide range of intermediate forms, some of which resemble one or other 

 parent very closely, and are usually recognised by the fact that they are completely sterile 

 with persistent spikelets on the mature culms. 



Much taxonomic confusion has been caused by the existence of this hybrid, which 

 has not been given adequate treatment by previous authors, and will therefore be discussed 

 here. 



The hybrid was first recognised by Townsend (1850, 1853) (see also 1904) who, 

 in a paper read before the Botanical Society of Edinburgh, in 1846 suggested for it the 

 name G. hybrida. Subsequently Townsend (1850) decided on the name G. X pedicellata 

 and as such read an account of the plant before the Botanical Society of Edinburgh on 

 13 December 1849, which was published in 1853. The species was cited from several 

 places in Cambridgeshire (v.c. 29), and at Dovedale, near Blockley (v.c. 57). The following 

 is a description of a specimen in Herb. Babington at Cambridge written by Townsend. 



" Perennial, with a stout, creeping rhizome. Culm ascending, stout, darkened 

 at the nodes, smooth, leaf-sheaths with cross-nerves, striate, glabrous and smooth; 

 ligule membranous, whitish, up to 10 mm. long; blades up to 15 cm. or more long, up 

 to 9 mm. wide, obtuse or more often acute, nerved on the upper surface, with the nerves 

 minutely scaberulous, smooth beneath. 



Panicle loose, ovate to pyramidal, up to 30 cm. long; axis smooth; branches 4 to 2 

 nate, filiform, gradually decreasing in length upwards, the longer at the base up to 11 cm. 

 long and this branched with the longer branchlets bearing 2-3 spikelets, the other basal 

 branches unequal in length, simple and from 4 to 1 spiculate, branches spreading, the 

 lower horizontally; pedicels 5-10 mm., spikelets more or less adpressed to the branches 

 and branchlets, linear - oblong 15 to 20 mm. long, 9 to 13 flowered, pale green. 



Glumes scarious, one nerved, obtuse; lower about 2 mm. long; upper 3 to 3-5 mm. 

 Lemmas at first imbricate, at length unrolled and exceeding the rachilla, thinly coriaceous 

 (dry) except for the hyaline tip, prominently 7-nerved and depressed between the nerves 

 when dry, broadly elliptic oblong and very obtuse when flattened 4 to 5 mm. long, 

 minutely scaberulous on the nerves and smooth or nearly smooth between them (Italics mine, 

 M.B.). 



" Paleas very slightly shorter than the lemmas, oblanceolate - elliptic minutely 

 2 - toothed, keels relatively broad-winged up to the apex, wings widest above the middle. 

 Anthers 1-1-2 mm., with imperfect pollen." 



Townsend's herbarium was presented to the South London Botanical Institute, 

 but the type specimen of G. X pedicellata is not there. Several of Townsend's specimens 



