1899] STUDIES ON REDUCTION IN PLANTS 13 
and irregularly oblong. There. is considerable variation in size. 
They are very characteristic in form and structure. A few 
show narrow slit-like openings in the middle. A few of this 
form become very short and the opening then is somewhat 
rounded, so that a ring form is the result. This is comparatively 
rare. Others are divided at the ends, some showing a slight 
emargination, while some are more or less deeply forked, show- 
ing a tendency to form XandY figures. Combinations of these 
two types are also found, so that there will be an opening in the 
middle, while the ends are somewhat forked, or one end is deeply 
divided. More characteristic, however, is the tendency to a 
differentiation in the density of the chromatin, which is especially 
marked when the chromatin is not stained too deeply. This dif- 
ferentiation of the chromatin is of the same kind as that mani- 
fested in the spirem and shows the paired condition of the 
chromosomes. The ground substance of the chromosome in 
these cases shows uniformly a paler tint and is translucent, while 
the denser masses of chromatin stain very dark and are more or 
less opaque. These chromatin masses are paired just as they are 
in the spirem, and probably result from the division of single 
chromatin masses in the early fission of the spirem. When these 
chromatin masses are well marked in the chromosomes, they 
seem to be uniformly of the same number so far as observed. 
There are often found well marked pairs of chromatin masses, 
each mass lying in the edge of the bar, four at the ends (two at 
each end) and four near the middle (two on each side near the 
middle), making eight in all. 
Very frequently the edge of the bar is undulate, the prom- 
inences on the edges occurring at the location of the chromatin 
masses. The apparent uniformity in the number of these 
chromatin masses in the chromosomes is, perhaps, of consider- 
able interest, and one is led to inquire whether they represent the 
units of the chromosomes or whether each one is a member of a 
tetrad group. If the latter is the true interpretation, then there 
would be in each chromosome of Trillium grandifiorum two 
united tetrads, or the chromosomes in the prophase of the 
