33 
To the second sub-group belong A. repens (L.) Beauv. and A. pungens 
(Pers.) R, & S. These are introduced from Europe, the former being 
found from the Atlantic to the Pacific, and the latter confined to sandy 
beaches on the coast of the Maritime Provinces. They do not form as 
dense mats as does A. Smithii, but grow either in rather open colonies or in 
small clumps with a few flowering stalks to each. The axis of the sterile 
shoots is more or less elongated, and the leaves, which on the living plant 
generally are spreading, in some cases almost horizontally, are fairly evenly 
distributed from near the base of the axis to the top. They are, so to 
speak, racemiformly arranged (Plate I, figure 1). The empty glumes are 
strongly nerved, generally acute or acuminate in A. repens, bluntish in A. 
pungens. 
The third sub-group includes A. dasystachyum (Hook.) Scribn., a very 
diversified species. The forms intergrade in a perplexing manner, the 
intergradations strikingly illustrating the taxonomic unimportance of such 
characters as presence or absence of pubescence and awn on the lemma. 
Series of forms may be found, in surprisingly small areas, with densely 
hirsute to glabrous lemma, and with perfectly awnless to long-awned lemma. 
Thus, to mention one example, numerous combinations of various degrees 
of pubescence and of length of awn have been observed by the writer at 
Calgary, Alberta, in an area only a few square rods in size. The forms of 
this sub-group either form small mats or clumps of varying density. This 
variation in habit cannot be taken as an indication of specific differences. 
It may very well be, and likely is, on a par with the variation in habit 
encountered in Festuca rubra L. and which, as has been experimentally 
proved, often is due to variation in the character of the soil. Thus a 
number of forms of Festuca rubra were, years ago, collected by the writer 
on a sandy and gravelly beach at Victoria, Vancouver island, B.C., where 
they were growing in extended open stands. When propagated in com- 
paratively rich, well-cultivated soil at the Central Experimental Farm, 
Ottawa, Out., they developed into dense, well-defined, tuft-like bunches, so 
dense, indeed, that had it not been for the characteristic basal sheaths, 
they might easily be taken for forms of F. ovina L. 
The forms of the A. dasystachyum sub-group all have numerous sterile 
shoots built after the fashion of those of A. Smithii (Plate II, figure 1). 
They differ from the latter by having strongly nerved, acute, empty glumes 
which generally are from one-half to two-thirds the length of the lowest 
lemmas. 
The second group of the second main group includes densely tufted 
forms without creeping rootstocks. Here belong A. spicatum (Pursh) 
Scribn. & Sm. and A. inerme (Scribn. & Sm.) Rydb. In these, the axis 
of the sterile shoots is elongated and their leaves arranged racemiformly as 
in A. repens (Plate II, figure 2). The empty glumes are strongly nerved 
and in shape and length much like those of the A. dasystachyum group. 
In herbaria, forms of the latter group with glabrous lemma may, therefore, 
easily be confused with forms of the A. spicatum group if the specimens 
are poorly and incompletely collected. In the field, however, the habit of 
growth and the construction of the sterile shoots make confusion impos- 
sible. To the writer’s opinion A. inerrne is a mere unawned variety of A. 
spicatum. 
35922— 3J 
