916 
Journal of Agricultural Research 
Yol. XXX, No. 10 
indicated a single-factor difference 
between intermedium and 6 -rowed. 
Crosses of deficiens with 2 -rowed also 
showed a one-factor difference. 
Arlington awnless, classed by Wig- 
gans (26) as an intermedium, has been 
used by Saunders and Moe (18) as the 
female parent in a series of crosses. 
Arlington awnless X 6 -rowed bearded 
produced in Fi an intermediate type 
and in F 2 Arlington awnless, intermedi¬ 
ate, and 6 -rowed in a 1 : 2 : 1 ratio. 
Arlington awnless X 2 -rowed hooded 
produced in Fi a type resembling the 
female parent for fertility of the 
florets in the lateral spikelets but with 
the florets of the medians hooded. In 
the second generation the plants were 
grouped into four classes: I. 2 -rowed 
hooded; II. 6 -rowed, or approximately 
6 -rowed, without awns on the florets of 
lateral spikelets but with hoods on 
the florets of the median spikelets; 
III. 2-rowed bearded with occasionally 
a few kernels present in the lateral 
rows; and IV. Arlington awnless type. 
Plants from Class I bred true in F 3 for 
the 2 -rowed condition but some families 
segregated for hoods and awns. The 
plants from Class II differed in the 
results they gave in F 3 . Plants from 
Class III bred true for 2 -rowed bearded. 
Plants from Class IV were of two types, 
one bred true for the Arlington awnless 
type and the other gave Arlington 
awnless, 2 -rowed bearded, and inter¬ 
mediates. 
Awns and hoods. —According to 
Hayes and Garber (8), Tschermak ob¬ 
tained results showing a single-factor 
difference between awnless and hooded 
forms, with the awnless character dom¬ 
inant. In crosses of hooded varieties 
with awned varieties the hooded con¬ 
dition tended to be dominant, the ratio 
in F 2 being 3 hooded to 1 awned. The 
same results were obtained by Thatcher. 
In a scries of studies, Ubisch (28, 24) 
found an interesting relation between 
the hooded and awned characters. He 
considered a spike as long-awned when 
the awns were 6 cm. long or longer, and 
as short-awned when the awns were 
shorter than 6 cm. 
In later studies Ubisch (24) crossed 
awnless varieties with short-awned and 
in some cases obtained an awnless Fi 
type and in others a hooded Fi type. 
The results of the later generations 
have not been published. 
Articulation of rachis. —In some 
varieties of barley, particularly of the 
species Hordeum spontaneum, the rachis 
is articulate and the heads shatter 
quite easily. A one-factor difference 
between the articulate and the 11011 - 
articulate rachis was found by Schie- 
mann (21) in a cross in which H. 
spontaneum was used as one of the 
parents. Articulate types were ob¬ 
tained by Ubisch (28, v. 17) from 
crosses in which both the parent 
varieties were nonarticulate. The ratio 
of plants with articulate rachis to 
plants with nonarticulate rachis was 
approximately 9:7 in the F 2 generation. 
I 11 one cross the articulation was as 
pronounced as that of H. spontaneum, 
but in the other crosses was less pro¬ 
nounced. 
Dense and lax spike. —The terms 
dense and lax, as used here, refer to the 
length of the rachis internode. In 
dense varieties the internode is short, 
and in the lax varieties long. The 
terms lax and dense are relative, since 
homozj^gous forms apparently exist (6) 
with small gradations all the way from 
very dense to very lax. Several factors 
for density of the rachis were shown to 
be present in one cross, while other 
crosses differed apparently by only a 
single differential factor. Ubisch (23, 
v. 17) considered the forms with inter¬ 
node length of 3.5 mm. or greater as 
lax and with inter node of less than 3.5 
mm. as dense. With this grouping he 
obtained one main factor difference for 
dense versus lax. He assumed other 
factors with lesser effect were present 
in some crosses. 
Rough awn and smooth awn. —In 
some varieties of barley the awn has 
saAvlike teeth from base to tip. Various 
types of teeth were described by 
Vavilov (25), who made a careful mor¬ 
phological study of the teeth on a 
number of varieties. He also found 
various degrees of smoothness but 
failed to find any \ T arieties Avhich lack 
the teeth on the entire length of the 
awn. Crosses of rougli-awned with 
smooth-awned varieties Avere first 
studied by Harlan (5), avIio found a 
single factor difference between rough 
aAvn and smooth with rough dominant. 
Vavilov obtained smooth-aAvned types 
in the F 2 generation from crosses in 
which both the parent varieties Avere 
rough-aAvned. In crosses of rough- 
awned A^arieties with smooth the Fi 
plants Avere always rougli-awned. 
Crosses Avere made between rougli- 
awned and smooth-awned varieties at 
the Minnesota Experiment Station 
(10). The awns of the Fi plants were 
rough. In the F 2 generation the rario 
of rough to smooth awned plants was 
approximately 3:1. Third-generation 
lines from F 2 plants classed as rougli- 
awned varied in their breeding be¬ 
havior. Some lines produced only 
rough-aAvned plants, some produced 
both rough and smooth, and others 
