10 BULLETIN" 622^ U. S. DEPAETMEN3? OF AGRICULTUKE. 
Minor distinctions useful hi describing suhvarieties: 
Outer glumes wide and narrow. 
Lemmas awnless and smooth awned as variations of awned. 
Kernels blue and purple as variations of white. 
Minor distinctions which are best used in describing agronomic biotypes (va- 
rieties founded on these are here disregarded except in the alphabetical list) : 
Only two outermost glumes expanded as a variation of wide outer 
glumes. 
Elevation of the hood on a short awn. 
Small awns arising from the hood. 
Flattened or twisted malformations of the awn. 
Short awns, except as noted in awnless. 
Character of the hairs of the rachilla. 
Scabrousness of the nerves of the lemma. 
Proliferation of the spike. 
' THE SPECIES OF CULTIVATED BARLEY. 
The species are of necessity based on fertility. This is predeter- 
mined essentially by the work already published, fertility having 
been used as a basis of separation since the earliest observations on 
barley. 
There is no such unity of opinion in fixing the number of species. 
Some favor placing all cultivated barleys under a single species, as 
did Jessen (1855) . To the writer, on the other hand, the group seems 
too broad to be so united. Clearness is better served by making the 
species a smaller unit. Certainly no group of wild plants of such 
variation is united under a single species, and there is abundant evi- 
dence in the behavior of hybrids that at least two parents were in- 
^,'olved in the production of the forms now domesticated. If fertility 
is to be subdivided there are three bases upon which it can be accom- 
plished. Many of the earlier writers recognize only two divisions. 
2 rowed and 6 rowed. Among the modern investigators Eegel takes 
this view. This has the disadvantage of throwing the hybrid inter- 
mediate forms with the 6-rowed, when physiologically th3 indications 
are that the}^ are nearer the 2-roAved forms. The only way to avoid 
this complication is to recognize the hybrid species intermedium of 
Kornicke. We then have three grades of fertility. There are actu- 
ally, however, four conditions. In one group of Abyssinian 2 -rowed 
barley the lateral spikelets are rudimentary. Although this distinc- 
tion is not as sharp as the others and needs a larger collection of 
Abyssinian material to determine its status clearly, there is no justifi- 
cation at present for rejecting the character. If three conditions are 
recognized, the fourth should be, which means the acceptance of the 
division de-flciens as equal in value to distichon and vulgare^ as sug- 
gested by Voss (1885). 
