1906] LEWIS—DEVELOPMENT OF RICCIA 123 
has ever been observed (fig. 34). When the nucleus is preparing 
for division, the chromatin leaves the linin network and collects into 
several bodies which soon move together to form one irregular mass. 
I regard this as the synapsis stage. Such bodies of chromatin have 
been found often and in cells which seemed to be well fixed so it 
Seems to represent a stage in the preparation for division and not 
to be a result of shrinkage as has been suggested by certain authors 
for other plant cells in which the same condition has been ‘observed. 
The body of chromatin occupies a position at one side of the nucleus, 
and the rather large nuclear cavity appears hyaline. There can 
be little doubt that the body described by CAMPBELL (3) as a nucleolus 
is really the’entire mass of chromatin in the synapsis stage. 
From this mass of chromatin a short thread develops which later 
Segments to produce the chromosomes (jig. 35). The small amount 
of chromatin present here makes the details very difficult to deter- 
mine. The four chromosomes, easily counted here as well as in 
the nuclear plate and on the way to the poles, are very small and 
appear almost spherical when on the spindle although they are 
short, thick, curved rods. 
The development of the spindle is not easily observed. Divisions 
take place almost simultaneously in all the cells of a sporogonium 
and the changes are very rapid. By far the commonest stage of 
division is that in which the chromosomes are in the nuclear plate 
(fig. 38). Neither centrosomes nor centrospheres occur in the spore 
mother-cells of either Riccia crystallina or Riccia natans. Around 
the nucleus preceding the formation of the spindle, there is an accu- 
mulation of material, apparently composed of fine fibres. The 
nucleus elongates, becoming somewhat elliptical but not sharp. 
pointed. The fibres about the nucleus do not give the appearance 
of centrospheres but are like the weft of kinoplasmic fibres described 
for certain pollen mother-cells (jigs. 36, 37). It has been impossible 
to find any nucleus which showed anything resembling a multipolar 
spindle. The poles of the spindle are probably determined by the 
elongation of the nucleus at an early stage in the spindle formation. 
The spindle is composed of very fine fibres, some of which extend 
from pole to pole, while others extend from the poles into the 
cytoplasm, reaching almost to the nuclear plate (fig. 38). The 
