In the present paper, following Mordvinkina (1936), 
binomials are restored to several of the taxa treated as 
subspecies by Thellung and Mal’tsev. This is scarcely 
justifiable in the case of A. strigosa sens. ampl. which 
seems to form a unified although poorly understood 
group. The division of subsection Denticulatae into the 
traditional four species is also questionable, it being prob- 
able that these oats comprise but one highly polymor- 
phous species. The present treatment follows common 
practice and is justifiable only for the sake of convenience. 
Key for the Determination of Species 
(largely after Mordvinkina, 1936) 
A. Lemma usually ending in two slender awns; plants mostly slen- 
der. (Subsect. Aristulatae Malz.) 
B. Glumes very unequal, upper almost twice as long as lower; 
2n= 14 (Series Inaequaliglumes Malz.) 
C. All florets of spikelet articulated, falling separately at ma- 
turity A. clauda 
C. Only lower floret articulated, spikelet falling as a unit 
A. pilosa 
B. Glumes equal or subequal 
D. Callus 5-10 mm. long, subulate; 2n=14 (Series Stipitatae 
Malz. ) 
E. All florets articulated A. longiglumis 
E. Only the lower floret articulated 
F. Callus+5 mm. long; glumes 25-30 mm. long 
A. ventricosa 
F. Callus to 10 mm. long; glumes to 40 mm. long 
A. Bruhnsiana 
D. Callus about 2 mm. long; 2n=14 or 28 (Series Eubarbatae 
Malz. ) 
G. Lemma ending in 2 awns and having 1 or 2 lateral teeth; 
glumes with 7-9 nerves 
H. Lemma with 1 lateral tooth (sometimes none); awns 
at tip of lemma usually exserted beyond the glumes 
I. Florets articulated, articulation-scar oblong-linear ; 
awns at tip of lemma to 12 mm. long A, hirtula 
I. Florets not articulated; awns 1-5 mm. long 
A, strigosa 
[ 269 ] 
