84 BOTANICAL GAZETTE [AUGUST 
synapsis is observed as a very definite linin thread in which deeply 
staining chromatic droplets (jig. 2) occur at intervals. This spirem 
is loosely wound in many convolutions through the nuclear area and 
shows no signs of fine anastomosing filaments uniting its parts. 
The nucleolus is not so conspicuous as in the preceding stages of 
synapsis. 
The spirem thread shortens and thickens and the chromatin 
granules become larger and less numerous (fig. 3). This process 
continues until the length of the whole thread is not more than that 
of the circumference of the nucleus, though it does not always occupy — 
a peripheral position. During the latter part of this shortening 
process, there is a crumpling of the thread and a crowding together 
of its chromatin granules. This condition is of comparatively short 
duration and it is at this point that I observed the first evidence of 
a double thread (figs. 4 and 5), 
The thread presently segments into eight chromosomes which lie 
scattered about in the nucleus in the form of a ring. These eight 
chromosomes are irregular in shape and frequently show with great 
clearness that they are not homogeneous masses, but made up of 
individual parts. I was at first in much doubt as to the number 
of parts, but subsequent study has convinced me that there are four, 
and that we are dealing here with tetrads. FARMER (7) shows 
very clearly by his figures of Fossombronia that he saw a similar 
arrangement. He says “sometimes four such aggregations could 
be seen in each chromosome, but the number was not sufficiently 
constant to afford very secure ground for theorizing.’ Howevel, 
he expresses the opinion that we have here a double longitudinal — 
split, in which the chromosomes are already prepared for the two 
succeeding divisions. : 
The tetrads are clearly shown in fig. 8. The appearance of 
several of the eight chromatic Masses suggests very strongly that they 
are made up of four parts, while the evidence presented by the on¢ — 
in the center is conclusive. Here there is present the ring form, — 
and the four elements of the tetrad are quite distinct. Several a7 
the forms assumed by the tetrads are illustrated in fig. 10, viz., CTOSSES 
Ys, Ts, and rings. The fourfold nature of the group is most cleat — 
in the ring form. In jig. 10a the four daughter chromosomes of the : 
