326 The Scottish Naturalist. 



Var. virescens, And. Chiefly remarkable for the short glumes 

 of the female spikelets. I have not seen Andersson's description, 

 and am indebted to Mr. Bennett for the name, which has been 

 confirmed by Dr. Almquist. 

 Deschampsia caespitosa, Beauv. 



The characters given for the varieties are so inconstant (while 

 different books give different varieties), that it is difficult to know 

 where to place some of our specimens. In the meantime I 

 arrange our forms thus : — 



a. genuina. The ordinary lowland form, with flowers more or 



less dark coloured. 



b. pallida , Koch (altissima,~L3.m.) A large, long-leafed, pale (yel- 



lowish white and green) flowered form, from shady places. 



c. hrevifolia, Parn. Short-leafed, flowers usually dark coloured. 

 High mountain specimens of this have smaller, closer panicles, 



with often 3 florets in the spikelet. I find that in all our forms 3 

 florets sometimes occur; but in these mountain plants their 

 occurrence is certainly more constant. In Mr. Cosmo Melville's 

 variety Voirlichensis of D. Jlexuosa, 3 florets in the spikelet are 

 always present ; so it would seem that the characters of the genus 

 must be amended. 

 Agropyron repens, Beauv. 



Under this I wish to call attention to a plant collected on Ben 

 Lawers by Mr Cosmo Melville, and which will probably be found 

 elsewhere if looked for. The specimen is in very bad condition 

 unfortunately, but comes nearest to A. repens. The spikelets are 

 apparently about 2-flowered ; the glumes and lower pales shortly 

 awned ; the ribs of empty glumes more scabrid ; and the ribs of 

 the flower glumes stronger and more scabrid. The most important 

 character, and one to which I have found nothing similar in any 

 of the numerous specimens of A. repens and A. caninum that I 

 have examined, is that the pale (that is, the upper pale of some 

 authors), has the two ribs more densely ciliate, and excurrent some 

 way below the tip of the pale, with two rough awns, which about 

 equal or surpass the ciliate triangular sub-acute and seemingly 

 entire tip of the pale. Can this be A. alpinum, Don M.S. ? 



Since this was written I have called Mr. Cosmo Melville's attention to the 

 plant, the result being that he has compared his own specimen (which is in 

 better condition) with A. violaancn Horn, and finds no essential difference. 

 A. violaceum is a Norwegian species and to it Nyman refers with a? Don's 

 alpimtm. 



