946 Davis.—A Comparison of the Reduction Divisions of 
of mid-synopsis (Fig. 9), becomes much more conspicuous in the post- 
synaptic stages leading to the development of the spireme. Thus with 
the loosening of the contracted coils of the synaptic knot there emerges 
a much more clearly defined system of loops than can be recognized in the 
complicated reticulum which entered the synaptic knot (compare Figs. 10 
and 11 with Figs. 6 and 7). It is evident that the threads have become 
much thicker and that the thread system has grown much shorter, and the 
inference is natural that these developments are the result of the process of 
contraction so characteristic of synapsis. This history is the same as that 
for biennis and grandiflora. 
Gates (’ 08 ) interprets the presence of parallel threads in post-synaptic 
stages of rubrinervis as indicating a splitting of the spireme, which is later 
closed by the union of the two threads. The writer, from his studies on 
biennis and Lamarckiana, cannot agree with this view, for although parallel 
threads may frequently be observed, there seems to him no evidence that 
this condition is other than such an arrangement as is to be expected when 
a complex system of threads has been gathered together by such a process 
as that of synaptic contraction. Since parallel threads may be observed 
in stages just previous to mid-synapsis it is but natural to expect that 
somewhat similar relations should be found in stages following the 
loosening of the synaptic knot. These threads in Oenothera are so delicate 
and their arrangement is so intricate that the establishment of such an 
interpretation as that of Gates is surrounded with great technical 
difficulties. 
As in the study of biennis , the term ‘ synapsis ’ is reserved for this period 
of a first contraction of the synaptic material. A type of contraction that 
is frequently conspicuous at a much later period during the formation of the 
chromosomes (the ‘ second contraction ’) will be described in that con¬ 
nexion. The relation and significance of the two phenomena will be 
further considered in the ‘ Cytological Discussion ’. 
The Formation of the Chromosomes. Following the period of 
synapsis, which is of relatively long duration, there begins as described 
above a gradual loosening of the synaptic knot and a shortening and 
thickening of the thread system. Further developments now follow rapidly. 
The threads become more prominent (Fig. 13) and begin to take on the 
thickness and appearance of the spireme which is soon to be developed. 
During this process the complicated coiled arrangement, which because 
of the anastomoses exhibited the structure »of a network, becomes more 
simple, and the threads may be followed for considerable distances in the 
looser convolutions. The threads remain, nevertheless, for a long time 
so complexly looped that their outline cannot be accurately traced through¬ 
out their entire length. 
The process of shortening and thickening proceeds until an undoubted 
