ent from maize chromosomes of comparable length. 
The pachytene chromosomes of this tropical form of 
Tripsacum australe, and those of the tropical form of 7. 
lavum, as well, are, for unknown reasons, more hetero- 
pycnotic than those of maize, especially along the short 
arms of the nine short chromosomes. 
Chromosome behavior was regular during all the stages 
of meiotic division. At diakinesis most of the homologues 
paired lengthwise, with occasionally one or two bivalents 
associated end-to-end. Univalents, however, were rarely 
found, 
Discussion 
One of the possibilities that may account for the origin 
of polyploidy in plants is by interspecific hybridization, 
The characteristics of the triploid form of Tripsacum 
lavum reported in this paper suggest that it probably 
originated in this manner. The evidence includes the 
following observations: (1) the pachytene chromosomes 
are heterozygous; (2) the chromosome behavior is ex- 
tremely irregular at meiosis; (3) viable seeds are rarely 
produced; and (4) more univalents than trivalents are 
found at diakinesis. 
In contrast to a triploid maize which Randolph and 
McClintock (1926) suggested had originated by the mat- 
ing of two gametes with the chromosome number of one 
of them became doubled in premeiotic division, the pres- 
ent triploid form of Tripsacum lavum probably had a 
tetraploid 7. davwmn as one of its ancestors and an un- 
known diploid as the second parent. According to the 
observations of the pachytene chromosomes, these hy po- 
thetical parental species were, at least, different in chro- 
mosomes 2, 3,5. 14 and 17. As stated in the foregoing 
section, these chromosomes were consistently found het- 
erozygous for knobs and large chromomeres, and in ad- 
dition, they often failed to associate regularly. 
[ 102 ] 
