1914] ELKINS—MATURATION IN SMILAX 4I 
and others). Mortrier (29) finds gemini in Tradescantia having 
double paired segments. This doubling is supposedly due to the 
reappearance of a split in the spireme which is believed to be a 
genuine homotypic fission and hence comparable to the split found 
in the paired segments of gemini in Smilax. 
Mode of reduction 
The mode of reduction in Smilax herbacea is essentially para- 
syndetic, though the procedure in the prophase seems to depart 
from the method described by Gr&GorrE (17) for parasyndesis. 
According to GREGOIRE, the chromatin in the prophase takes the 
form of thin paired filaments (leptonema) which fuse (zygonema), 
shorten and thicken (pachynema), and again separate (strepsinema). 
In Smilax the chromatin in the prophase is distributed in granules, 
which are frequently seen in pairs. That which takes place between 
this condition and the spireme is obscured by synapsis. The con- 
struction and relative arrangement of the chromosomes in the 
spireme can be inferred only from subsequent behavior. After 
synapsis, two longitudinal splits occur; the first appears early in the 
spireme and can be traced through strepsinema to diakinesis; the 
second split shows sometimes in the univalent halves of the gemini 
at diakinesis, but more often not until the metaphase. From this 
we may infer that the first doubling is a separation of previously 
paired elements and that the chromosomes or chromatic bodies are 
placed side by side in the spireme. The second doubling is a 
genuine fission. 
Discussion 
PERSISTENCE OF CHROMOSOMES.—Much of the cytological work 
of recent years has brought forward directly or indirectly the 
question of the persistence of chromosomes or of some smaller unit. 
The rediscovery of the work of MENDEL has given added impetus 
to the hope of finding a physical basis for heredity or the unit 
characters of MENDEL which, according to experiment, seem to 
pass from generation to generation inviolate. Opinions concern- 
ing the existence of this cytological unit necessarily vary. Some 
favor the idea of persistent chromosomes; some maintain that 
