Feb. 23, 1924 
Teosinte Maize Hybrids 
575 
of teosinte. The rudiments of intermediate structures are found, but 
they remain abortive and sterile. 
It is a moot question whether the simple pistillate inflorescence of 
maize represents the loss of a more specialized branched condition or has 
come about through a still further specialization of a branched inflor¬ 
escence in which the branches have become united. The ramose varia¬ 
tion is the principal evidence in support of the former view, and it is 
clear that the branched condition of ramose maize had nothing in com¬ 
mon with the branched inflorescence of teosinte. The incompatibility 
of the ramose form of branched inflorescence in combination with the 
other characters of the pistillate spike of teosinte seems all the more 
remarkable in view of the general tendency toward branching mani¬ 
fested throughout the teosinte plant, and including both vegetative and 
floral organs. 
Although it might have been expected that the ramose variation would 
be correlated with maize characters and that few nonramose inflores¬ 
cences with many rows of alicoles would be obtained, the contrary proved 
to be the case. Thus, while the number of F 2 individuals grown from the 
ramose hybrid was less than that of any of the other hybrids, the parental 
types of inflorescences characterizing the normal forms of the parent 
species were recovered more nearly than in other hybrids. The best 
approximation to a normal unbranched ear was found in the F 2 of the 
ramose hybrid, and an inflorescence typical of Durango teosinte also was 
obtained ,though no spikes were found in either the crinkly or ramose 
hybrids having seeds shaped like those of Florida teosinte (4). (See 
PI. 7, D.) 
Not only were maizelike unbranched pistillate inflorescences obtained, 
but even in the relatively small group of ramose plants branched ears 
were found closely approximating the ramose ears of maize. In none 
of the cases of maizelike ears, however, were the inflorescences as large 
as the normal ears of the variety from which the ramose parent was 
obtained, the longest ear being less than 15 cm. in length. 
The group of ramose pistillate infloresences embraced no other ances¬ 
tral characters. Thus, as in maize, when fertile pistillate spikelets were 
produced, these always had the hardened outer glumes and reduced 
bractlike inner glumes characteristic of the spikelets of a maize ear. 
There was a tendency for the inflorescences to be staminate, but this 
largely is a measure of sterility, which was so pronounced in the ramose 
plants. Among the ramose inflorescences that approached the two- 
rowed teosinte spike, the specialized rachis, characteristic of the pistillate 
inflorescence of teosinte, is entirely lacking, the inflorescence resembling 
completely a branch of the staminate panicle, with the distance between 
the articulations greater than is the case in normal pistillate inflorescences. 
CORRELATION OF THE RAMOSE VARIATION WITH CHARACTERS THAT 
DISTINGUISH MAIZE FROM TEOSINTE 
When the biserial correlations with type of plant are examined there 
are found to be only eight coefficients exceeding three times their error. 
These coefficients are given in Table VII. The biserial method could 
not be applied to the characters that measured the degree of ramoseness, 
since the distribution of plants with respect to these characters was 
bimodal. Of the eight coefficients, five at first sight appear to be dis- 
herences, while only three, days pollen to silk, height, and length of 
