THE IDENTIFICATION OF VARIETIES OF BAELEY. 13 
uncommon, and deficient barleys which show intergradations toward 
the normal 2-rowed are rare indeed. 
In the 6-rowed barleys all awnless forms are thrown into inter- 
Tnedium. The awnless character is most probably of hybrid origin, 
and in all 6-rowed awnless varieties now known there is a tendency 
in some strains to produce short awns on the central spikelets when 
grown under especially favorable conditions, indicating a direct rela- 
tionship with intermedium: It may be found necessary to include 
awnless sorts under Hovdeurrb vulgare as well, but at present it seems 
better to list these forms with the intermedium. 
Of the varieties and subvarieties which follow, a large number are 
of hybrid origin. The species intermedium probably consists entirely 
of hybrid varieties. In the beginning it was attempted to separate 
the known hybrid varieties from those occurring naturally. How- 
ever, there seems no good reason for such a distinction, for many of 
the varieties which are not known to be hybrid probably have arisen 
from accidental crossing. 
For example, the variety Jiorsfordianum was named for a Mr. 
Horsford in Vermont, who crossed a Nepal (6-rowed hooded naked) 
on a 6-rowed bearded hulled and obtained a 6-rowed hooded hulled. 
This variety is grown extensively in the United States and is abso- 
lutely constant. It is inconceivable that all barley of this descrip- 
tion in this country and elsewhere came from Mr. Horsford's cross. 
The Nepal has been grown in India alongside of hulled varieties for 
centuries, and the cross must have occurred time and again naturally, 
despite the fact that barley is one of the most closely fertilized of 
plants. It is a strange coincidence that the only natural hybrid that 
has occurred in the nursery in the six years of this investigation was 
between the Nepal and a 6-rowed bearded hulled variety, from the 
progeny of which a Horsford type has been isolated. 
In describing varieties, fertility having been used as a specific dis- 
tinction, there remain hulled or naked kernels, awned (including 
awnless) or hooded lemmas in the central spikelet, and white (includ- 
ing blue and purple) or black color as major distinctions. The first 
two of these characters, with the variation in width of the outer 
glumes later spoken of in the discussion of subvarieties, are shown 
in Plate III. The regularity of the gradation in differences be- 
tween the varieties in each group is sufficient to excite the suspi- 
cion of the taxonomist, because such uniformity is not common in 
nature. Ordinarily, variations do not occur in a closely graded series. 
Neither did they in barley at first, but continued crossing, natural and 
artificial, has slowly filled the gaps. 
Hordeum vulgare trifurcatum, for example, at first stood off 
clearly as the trifurcate naked 6-rowed barley. Then horsfordianum. 
was produced and there were two trifurcate 6-rowed sorts, one hulled 
