Jan. 1,1925 
A Study of Variability in the Burt Oat 
35 
Kernels described in 1920 as subclass 
abundant long had a dense growth of 
long basal hairs which sometimes 
almost surrounded the base of the 
lemma (PL 2, A). The basal hairs 
described as abundant midlength (PL 
2, B) were not so long as those in the 
abundant long subclass and often they 
were not produced so abundantly. 
Kernels described as having few basal 
hairs usually had only a few short 
hairs around the base of the kernel 
(PL 2, C), although some kernels bear¬ 
ing a very few long hairs were included. 
Kernels which bore long hairs usually 
had them abundantly developed. Ker¬ 
nels bearing no visible hairs were classed 
as absent (Pl. 1, B, C, E). 
STUDY IN 1920 
The data on basal hairs from both 
stations in 1920, presented in Table V, 
clearly indicate that all of the three 
parental types segregated to a consider¬ 
able degree. The abundant class was 
a little more constant in breeding be¬ 
havior than either of the others. The 
class termed few was the least constant. 
In general, however, each parental class 
produced progenies having a slightly 
larger number of kernels of its own class 
than of either of the other classes. 
An examination of the data for each 
strain shows that in only five of the 
eight strains grown at Akron were any 
of the parental kernels described as 
having abundant hairs. If the per¬ 
centages of kernels classed as abundant 
midlength and abundant long pro¬ 
duced from the parental kernels in the 
abundant class are added together for 
all strains grown at Akron, nearly 85 
per cent of the progeny kernels are in 
the two abundant subclasses. The 
percentages range from 75 per cent in 
Kansas No. 6076 to 93.5 per cent in 
Kansas No. 6090. In all strains grown 
at Akron in 1920 the parental kernels 
described as having abundant basal 
hairs produced progenies in which 
practically no kernels were classed as 
having basal hairs absent, although 
9.2 per cent of such kernels occurred 
in Kansas No. 5211. In Kansas No. 
5020 only two kernels and in Kansas 
No. 6076 only one kernel having no 
basal hairs occurred in progenies of 
parental kernels having abundant hairs. 
As a rule the parental kernels classed 
as having abundant basal hairs were 
more nearly constant in breeding be¬ 
havior than either of the other two 
classes. These data show clearly how 
few kernels genetically having long 
hairs were classed as having few hairs. 
In many of the strains the kernels 
classed as having few basal hairs pro¬ 
duced a majority of kernels of the 
abundant hair class. Of all of the 
strains at both stations parental kernels 
with few hairs produced only 2.5 per 
cent of progeny kernels with hairs 
classed as abundant long and only 
5.1 per cent classed as absent, while 
44.9 per cent were classed as abundant 
midlength, and 47.5 per cent as few. 
The parental kernels classed as having 
no basal hairs produced 6.8 per cent 
with hairs described as abundant mid¬ 
length, and only 0.3 per cent with abun¬ 
dant long hairs, but 36.4 per cent with 
few, and 56.5 per cent with hairs absent. 
Of all of the eight strains only 36.4 
per cent of the progeny of parental 
kernels having absent basal hairs were 
classed as having few hairs. 
The summary shows that all parental 
classes produced some kernels in all of 
the classes of basal hairs. Those 
parental kernels classed as having 
abundant hairs bred true in 76 per 
cent of the cases, if the abundant 
long and midlength subclasses are con¬ 
sidered together. Parental kernels 
classed as having few hairs produced 
progenies consisting very largely of 
the abundant midlength and few 
classes. The parental kernels classed 
as having hairs absent produced prog¬ 
eny of which only 56.5 per cent were so 
described, while 43.5 per cent bore 
basal hairs. 
STUDY IN 1921 
The 1921 data on inheritance of basal 
hairs are presented in Table VI. No 
kernels described as having abundant 
long basal hairs were sown at Manhattan. 
At Akron these kernels produced proge¬ 
ny of which 95 per cent were like the 
parents and 5 per cent had abundant 
midlength hairs. This would indicate 
that the abundant long class of basal 
hairs breeds as a recessive. Kernels de¬ 
scribed as having abundant midlength 
basal hairs produced progenies of which 
53.3 per cent were of that class, about 
30 per cent were abundant long, 16 
per cent few, and less than 1 per cent 
absent. 
Parental kernels described as having 
few basal hairs produced progeny of 
which the basal hairs on less than 1 
per cent were classed as abundant long, 
on about 46 per cent as abundant mid¬ 
length, on 46 per cent as few, and on 
about 7 per cent as absent. 
The progeny of parental kernels de¬ 
scribed as having basal hairs absent 
were distributed about as follows: 62 
per cent absent, 36 per cent few, 2 
per cent abundant midlength, and less 
than 1 per cent abundant long. 
