16 BULLETIN 925, U. S. DEPARTMENT OF AGRICULTURE. 
• 
chjtic character with a partial failure of dominance or they may be 
considered as instances where the segregation has been incomplete, 
In either event the well-defined difference between these brachvtic 
internodes and the adjoining normal nodes emphasizes the indi- 
viduality of the node and internode as a metamer. If these cases are 
looked upon as partial failures of dominance, it is significant that 
instead of a general shortening of internode length throughout the 
entire plant single internodes only are affected and become abruptly 
and decidedly shorter than their mates. This fact indicates that no 
general lengthening of the internodes is to be expected as the result 
of contamination by repeated " back crosses " on plants of normal 
stature. 
These cases would seem to offer some support for the hypothesis 
of somatic segregation or perhaps bud variation. If such an ex- 
planation is adopted, a question arises whether an ear borne on a 
node subtended by a brachytic internode would behave in heredit} T as 
a true brachytic. Xo such cases have arisen as yet, and there would 
be some difficulty in distinguishing them unless the brachysm was 
pronounced, as in the case shown in Plates IX and X. 
The internode borne on the same node with the ear is almost in- 
variably shorter than the internodes immediately above and below, 
so that to be detected brachysm of the internode at this point must 
of necessity be pronounced. 
Although an ear borne on a brachytic internode might be ger- 
minally brachytic, it of necessity must be fertilized by male gametes 
from an inflorescence produced from a normal node. The resulting 
seeds, therefore, would give rise either to plants of normal stature 
only, or, if the parental plant had been heterozygous for brachysm, 
one-half of the progeny could be brachytic. The point is of interest 
in connection with the sudden appearance of braclrysm in the two 
Chinese hybrids. 
The brachytic segregates in the F 2 of Boone-brachytic hybrids 
had a mean height in decimeters of 10.42 ±0.112, with a coefficient 
of variability of 11.3 ±0.79 and a range of 7 to 15 decimeters. The 
coefficient of variability of the brachytic segregates compared with 
that of the 11 brachytic progenies shows that there was no signifi- 
cant increase in variation in the segregates, the increase being only 
1.2±0.S7. A somewhat larger increase is obtained if the coefficient 
of variability of the brachytic segregates is compared with that of 
the brachytic progeny grown in 1918; the difference then becomes 
4.2±1.13. or 3.7 times the probable error, but since these progenies 
were not grown the same year the propriety of comparing their 
variability is questionable. While the average height of the bra- 
chytic segregates was greater than that of brachytic plants directly 
descended from the original varietv, the difference of 1.61±0.87 is 
