1912] STEVENS—HETEROSTYLOUS PLANTS 293 
téne loops exist in the haploid number, and soon undergo a longi- 
tudinal division, which is really the separation of the parts which 
united in their formation. The halves thus separated often appear 
irregularly spread apart and crossed, the “strepsiténe” stage; 
and the “gemini” are formed by the shortening and thickening 
of these strepsiténe loops. Diakinesis (p. 232) is characterized 
by the presence of “chromosomes” in the reduced or haploid 
number, formed often of two rather independent branches. The 
first division separates the two branches of these “ chromosomes,” 
and the daughter chromosomes show during the anaphase of the 
first division a longitudinal split, which is sometimes visible in the 
two branches of the diakinetic “chromosomes.” After an inter- 
kinesis, more or less brief, marked by varying degrees of nuclear 
reconstruction, the daughter chromosomes of the first division 
reappear, and their longitudinal halves are separated in the second 
division. 
In Grécorre’s scheme, synapsis, when it occurs, is due to the 
contraction which accompanies the fusion, in pairs, of the lepto- 
téne threads to form the pachyténe loops. If this change takes 
place simultaneously throughout the nucleus, a crowding of the 
whole chromatin mass at one side is the result. 
If it is assumed, as seems probable, that in the buckwheat a 
fusion of thin filaments takes place in the presynaptic stages, then 
the series of loops characteristic of the early postsynaptic period 
represents a pachyténe stage. There are, however, considerably 
more than the haploid number of loops. This may mean either 
that the loops representing the chromosomes are long (GREGOIRE, 
p- 335) or that they are parts of a continuous spirem thread. These 
two explanations represent the two interpretations of the spirem 
condition, STRASBURGER and his school maintaining that the 
clivomatic mass comes out of synapsis as a pirem thread, 
Grécore holding that the so-called spirem is really a series of 
independent loops. On this point the buckwheat furnishes no 
evidence. It seems certain, however, that these loops become 
shortened and thickened and undergo a longitudinal split, forming 
the strepsiténe loops, from which the gemini are derived by a 
continued thickening. 
bas 
