422 BOTANICAL GAZETTE [DECEMBER 
of ALLEN (I, 2, 3, 4), GREGOIRE (35), BERGHs (7, 8, 9, 10), ROSEN- 
BERG (64, 65, 66), and others. 
The period of synapsis lasts for some time, after which the spirem 
becomes looser and distributed throughout the nuclear cavity. This 
thread, at first uniform in thickness, begins to split longitudinally 
(jigs. 152, 152a), the two parts lying close together side by side for 
long distances, but sometimes diverging at wide angles and then 
coming together again. The transverse segmentation of this double- 
spirem to form the chromosomes now takes place, although not 
simultaneously throughout the entire nucleus. The shape of these 
chromosome segments is very irregular when first formed. They 
may be bent or twisted like two Vs (fig. 153a, V), or two Ls (jig. 1534; 
L) placed one above the other, or crossing in the form of an X (7g. 
1534, X). The segments gradually shorten (jig. 154) until 20 short 
rod-shaped chromosomes, bivalent in nature, are formed (jig. 155): 
These are of course really the 40 sporophytic chromosomes now 
grouped in pairs. The nucleolus present in the resting nucleus of 
the tetraspore mother cell has not changed visibly up to this prophase 
stage, taking various positions, and during synapsis being surrounded 
by the tangled mass of threads. Thus the 40 chromosomes which 
entered the tetraspore mother cell now appear after synapsis, which 
is generally believed to be the period of chromosome reduction, as 
20 pairs. These pairs become arranged in an equatorial plate and 
the chromosomes of the pairs split longitudinally, so that a large 
number of chromosomes results (fig. 156), probably 80 in all, 
but so crowded that it is not possible to count them with absolute 
certainty. 
While these changes are taking place inside of the nucleus the 
kinoplasm accumulates in two opposite poles outside of the nuclear 
membrane. Each pole of the spindle which is formed (fig. 156) 18 
occupied by a deeply staining centrosome-like body, as in the case of 
the prophase of the other mitoses previously described. The form 
of this spindle in the tetraspore mother cell is quite different from 
other mitoses. Its longest axis runs from pole to pole instead of 
across the equatorial plate, as in the other nuclear figures. Besides, 
the two poles are less than 180° apart, which gives an asymmetric 
or somewhat bent form to the spindle when viewed from a certain 
