32 
Journal of Agricultural Research 
Vol. XXX, No. 1 
two stations. This probably was part¬ 
ly owing to the fact that the descrip¬ 
tions of the kernels grown at the two 
stations in 1920 were made by different 
persons. The percentage of basifrac¬ 
ture kernels in the progeny of parental 
kernels described as disjoining by disar¬ 
ticulation was much lower at Akron, 8 per 
cent, than at Manhattan, 21.7 per cent. 
In the progeny of parental kernels in 
which disjunction was described as by 
basifracture, 13.5 per cent of the kernels 
grown at Akron and 31.8 per cent of the 
kernels grown at Manhattan were so 
described. In the progenies of paren¬ 
tal kernels in which disjunction was 
described as by basifracture there were 
about 61 per cent by disarticulation in 
the Manhattan material and about 42 
per cent in the Akron material. 
The highest percentage of progeny 
kernels of which floret disjunction was 
described as by basifracture, produced 
from parental kernels so classed, was 
found in strain Kansas No. 5211 at 
Akron and in strain No. 5219 at Man¬ 
hattan. Kansas No. 5219 was not 
grown at Akron. The percentage of 
disjunction by disarticultion in the 
progeny of parental kernels in which 
disjunction was described as by basi¬ 
fracture was not far below 60 per cent 
in any of the strains at either station, 
with the exception of Kansas No. 6090, 
grown at Akron, and Kansas No. 5219, 
grown at Manhattan. In Kansas No. 
6052, grown at Akron, and Kansas 
Nos. 6052 and 6076, grown at Manhat¬ 
tan, less than 5 per cent of the progeny 
kernels of parental kernels in which 
floret disjunction was described as by 
basifracture were so classed. In each 
of these three cases, in about 90 per 
cent or more of the progeny of the basi¬ 
fracture parental type, disjunction was 
described as by disarticulation. The 
comparatively small number of indi¬ 
viduals of each of these strains, how¬ 
ever, makes it unwise to consider these 
results conclusive. 
An examination of the data for each 
of the strains at both stations indicates 
that of the disjunction types the lowest 
percentages of progeny described as 
disarticulating, from parental kernels 
so described, were produced in those 
strains which had the highest percent¬ 
ages of progeny disjunction by basi¬ 
fracture produced from parents of that 
type. In most of the strains grown at 
each station approximately two-thirds, 
and in many cases three-fourths or 
more, of the progeny from parental 
kernels disjoining by disarticulation 
were described as disarticulating. 
The results of the 1920 studies on 
floret disjunction clearly indicate that 
the form resulting from basifracture in 
all of the strains under observation was 
much less stable in breeding behavior 
than was the form resulting from dis¬ 
articulation. A comparatively small 
percentage of the parental kernels in 
which disjunction took place by basi¬ 
fracture produced progeny so described, 
while in most strains at both stations 
two-thirds or more of the progeny ker¬ 
nels from parental kernels in which 
disjunction was described as by disar¬ 
ticulation were of this type. 
STUDY IN 1921 
Floret disjunction in the progeny 
kernels grown in 1921 at Akron was 
described in the same way as in 1920. 
(Chinch bugs damaged the oat plants 
at Manhattan to such an extent that it 
is impossible to include data for the 
1921 crop at that place.) The results 
obtained in 1921, presented in Table 
IV, are in general agreement with those 
obtained in 1920. 
An examination of the data shows 
that a comparatively small percentage 
of the kernels in which disjunction was 
classed as by basifracture bred true. 
In each of the five strains where prog¬ 
enies were produced from kernels so 
described less than 20 per cent of the 
progeny kernels in any strain disjoined 
by basifracture. This indicates that in 
the Burt oat used in these experiments 
disjunction by basifracture is very un¬ 
stable in its breeding behavior, pro¬ 
ducing larger percentages of progeny in 
which disjunction takes place by dis¬ 
articulation and by heterofracture than 
by basifracture even in the second 
year of pedigree culture. 
The kernels in which disjunction 
was described as by heterofracture in 
1920 produced progenies in 1921 of 
which approximately 45 per cent of the 
kernels disjoined by heterofracture, 44 
per cent by disarticulation, and 11 
per cent by basifracture. It is possible 
that in describing the 1920 material a 
larger number of kernels of which 
disjunction was genetically by disarticu¬ 
lation than of those in which it was 
genetically by basifracture was classed 
as by heterofracture. 
The results obtained in the progenies 
from kernels of which disjunction was 
classed as by disarticulation in 1920 
indicate that disarticulation is com¬ 
paratively stable in breeding behavior. 
The progeny kernels produced in 1921 
from the kernels of which disjunction 
was by disarticulation showed about 
74 per cent disjoining by disarticula¬ 
tion, 22 per cent by heterofracture, and 
less than 5 per cent by basifracture. 
