RAMOSE INFLORESCENCE IN MAIZE. 17 
maize hybrids, found evidence that the number of rows was increased 
through a twisting of the rachis. Hackel's hypothesis of coalescence 
does not account for the arrangement of branches on the staminate 
and pistillate inflorescences of either ramose or normal plants. The 
Eamosa-Gordo hybrids furnish evidence that the present type of in- 
florescence developed through the reduction of branches which pre- 
sumably occurred subsequent to the twisting of the rachis. 
It is difficult, however, to account for the present complete differ- 
entiation of branching space and central spike in the staminate in- 
florescence on the theory of reduced branches, since it necessitates 
the assumption that the reduction ceased abruptly at about the mid- 
point of the staminate inflorescence, while it continued to the base of 
the pistillate inflorescence. 
The frequent appearance of branched but nonramose ears where 
the pistillate inflorescence is more nearly of the form of the common 
staminate inflorescence strongly suggests that the loss of the basal 
branches followed rather than preceded the separation of the sexes. 
If the view is adopted that the many-rowed spikes originated 
through the reduction of branches, it must be concluded that this 
reduction took place in two distinct periods. Thus, the upper two- 
thirds of the spike was developed through branch reduction, result- 
ing in an inflorescence with a central spike and basal branches, a 
condition which was stabilized in the staminate inflorescences, while 
in the pistillate inflorescences the remaining basal branches have been 
reduced to spikelets or possibly entirely aborted. With this idea 
of the reduction of branches in restricted areas, it becomes possible 
to explain the bifurcated or bear's-foot ears by assuming that the 
reduction in branches began somewhat below the apex at a point 
corresponding to the place where the staminate spikelets first open. 
The branches above this point were reduced later, and the bear's-foot 
ear is a variation entirely comparable with the development of basal 
branches. In this connection it may be more than a coincidence that 
the branched-ear variation in Pawnee appeared simultaneously with 
a distinct bifurcation at the apex representing a reversion to the 
condition suggested in this hypothesis that the central section of the 
branches were the first to be reduced. In other words, the reduction 
in branches first took place on a central section of the rachis, leav- 
ing the tip and base branched, though the apical branches, of course, 
were short. These two sections were reduced later, and from the 
frequency of their reappearance it may be not unreasonable to infer 
that the branches of the tip were the last to go. On the other hand, 
if the change in the form of the pistillate inflorescence was entirely 
similar to that of the staminate inflorescence, the order of the dis- 
appearance of the branches must be reversed, since the basal branches 
of the staminate inflorescence have yet to be reduced except in the 
