pachytene, the 18 chromosomes of a haploid set could 
be recognized. The length, arm ratio and spindle fiber 
attachment region of each chromosome were identified. 
There are probably three chromosomes (8, 5 and 8) hav- 
ing internal knobs on the long arm. The internal knob 
on the long arm of chromosome 8 is small and definitely 
heterozygous. The other two internal knobs are large and 
homozygous (Plate XX, figs. 1 and 2). Probably only 
chromosome 9 is knobless. The other 17 chromosomes 
have one or two knobs terminating one or both arms. 
However, there is a clear tendency to have the knobs 
terminating the long arms. Chromosomes 2, 8, 5, 14 and 
17 are heterozygous for knobs and prominent chromo- 
meres. In agreement with Longley’s report (1987) on 
Tripsacum floridanum, chromosome 16 has a nucleolar 
organizer on its long arm. This condition is different from 
that in maize. Furthermore, the secondary constriction 
of the nucleolar chromosome in 7". /awwm is not always 
well marked. 
Univalent chromosomes were frequently entangled 
with the bivalent chromosomes in a densely crowded 
mass. Whenever they were isolated, they demonstrated 
the spindle fiber attachment region and were readily iden- 
tifiable. It was also observed that the univalent chromo- 
somes sometimes formed a non-homologous association. 
Chromosome fragments and loops in the bivalent chro- 
mosomes were frequently found at pachytene. Fusions 
of the chromosome knobs and those of the spindle fiber 
attachment regions were constantly present. 
At diakinesis, eighteen bivalent chromosomes were 
almost never found ; the number of the univalent chromo- 
somes was always greater than eighteen. Multivalent 
chromosomes were also seen, but only with a low fre- 
quency. At metaphase I, many chromosomes often failed 
to congress in the equatorial plate; the number of lag- 
[ 100 | 
