20 
Journal of Agricultural Research 
Yol. XXX, No. 1 
ity apparently is evident only in 
mature or nearly mature specimens. 
In many fresh, immature specimens 
both of wild and cultivated oats the 
writers have observed no hollow or 
scar in the callus or the basal segment 
of the rachilla when these organs are 
forcibly disjoined. 
In the present study the different 
ways by which spikelet disarticulation 
takes place were classified as resulting 
from (1) abscission when the method 
of spikelet separation was that charac¬ 
teristic of the wild Avena fatua or A. 
sterilis, that is, resulting in a pronounced 
cavity or scar in the base of the lemma; 
(2) fracture when the method of 
separation was that most characteristic 
of the cultivated varieties of A. sativa, 
that is, resulting in a roughened tissue 
with no observable cavity or scar at 
the base of the lemma; and (3) semi¬ 
abscission when the method of separa¬ 
tion was to some extent intermediate 
between the two, apparently resulting 
partly from abscission and partly from 
fracture, and leaving only a slight and 
often poorly developed cavity or scar 
in the base of the lemma. These three 
kinds of disarticulation are shown in 
Plate 1, A to C. 
FLORET DISJUNCTION 
The florets of an oat spikelet are 
connected by the clavate segments of 
the rachilla. The articulation at the 
juncture of two segments is not very 
evident, owing to the structure of the 
spikelet. The base of the lemma of 
each floret is attached to the enlarged 
apex of the rachilla segment which 
bears it and entirely surrounds the 
slender base of the segment which 
bears the next higher floret. This 
latter segment, therefore, appears to 
grow out from the basal tissue of the 
inrolled lemma next below (PI. 1, D). 
The manner of separation of the florets 
from one another, here termed floret 
disjunction, differs in different species. 
The rachilla segments usually are 
narrowly clavate in shape and variously 
flattened, rounded, or furrowed. They 
usually are 1.5 to 3.5 mm. long, except 
in Avena nuda , where they are extremely 
elongated. In the wild species they 
often are clothed by stiff hairs. The 
number and character of these hairs 
are much used in differentiating be¬ 
tween wild and cultivated forms. 
In some species, as Avena fatua and 
its derivative A. sativa , disjunction 
of the second and additional florets 
usually takes place by disarticulation 
at the apex of the rachilla segment. 
In such cases the rachilla segment 
remains attached ventrallv at the base 
of the next lower floret of the spikelet. 
The attachment of the rachilla segment 
to the base of the second and subse¬ 
quent florets is very firm in A. sterilis 
and its derivative A. byzantina. Tho 
florets of the spikelet often remain 
attached together during threshing. 
When forcibly separated, the segment 
breaks near its base, as pointed out 
above. But in A. sativa the separa¬ 
tion by disarticulation at the base 
of the floret usually takes place very 
readily. 
In the present study it was observed 
that in some cases the segment breaks 
across near the middle or even splits 
irregularly lengthwise. 
Floret disjunction in this study is 
described in terms of the nature of the 
disjunction of the supporting rachilla 
segment. The term (1) disarticulation 
is used when the segment disarticulates 
from its floret, as in Avena fatua ; (2) 
the term basifracture when disjunction 
is by fracture at the base and it 
remains attached to the floret, as in 
A. sterilis; and (3) the term hetero¬ 
fracture when it fractures irregularly 
in the middle and the disjunction can 
not be classified definitely. Plate 1, 
D to F, shows these three different 
types of floret disjunction 
BASAL HAIRS 
The surface of the lemma may be 
either hairy or glabrous, characters 
much used in separating oat species. 
Most wild species of oats are character¬ 
ized by hairiness of the lemma, callus, 
and rachilla. The callus often bears 
more or less conspicuous bristles, 
usually termed basal hairs. The pres¬ 
ence of these hairs may be observed 
readily without magnification. 
The classification of basal hairs in 
the present paper conforms somewhat 
to that of Zade {156). He classed 
them as long, intermediate, and short. 
The writers have discussed them as 
(1) abundant long, (2) abundant mid¬ 
length, (3) few, disregarding length, 
and (4) absent. 
The abundant long class of the 
writers is practically identical with the 
long class of Zade. The abundant 
midlength class includes most of those 
he described as intermediate. In 
addition, the writers included in this 
second class a few kernels bearing 
hairs which though rather short were 
very abundant. Some of these kernels 
doubtless would have been put in 
Zade’s short class. The writers ’ “few” 
class was in reality a sort of catchall. 
It included all kernels that were only 
