350 
LESTER W. SHARP 
assumed, or, as is in all probability actually the case, by the formation of a 
median series of vacuoles and spaces which are for the most part if not 
entirely of prophasic origin. It may again be emphasized that in Trades- 
cantia, as in Vicia, the alveolar-reticulate condition of the telophase and 
the early prophase does not pass directly into the split spirem stage as 
above provisionally assumed; but rather, by a peculiar process in which 
most if not all of the old vacuoles and openings are lost, the reticulate 
chromosome takes the form of a single thread in which the split then 
develops anew. In these forms the chromosome split, whether entirely 
new or partially built of vacuoles of earlier origin, cannot be regarded as 
anything but a prophasic development. 
Method of chromosome splitting. Much interest centers about the exact 
manner in which the chromosomes undergo splitting because of its great 
importance in connection with current theories of the mechanism of heredity. 
The early view that the splitting of the chromosomes is primarily a division 
of a series of smaller units which it contains has been widely accepted, 
especially by those whose investigations have been concerned with the 
cytological aspects of the problem of inheritance. The hypothesis which 
postulates the existence of small units of inheritance, or genes, within the 
chromosomes has been of incalculable value in organizing the data of 
genetics, and new evidence constantly adds to the probability that such 
units are not purely conceptual ones. But it must be said that this evidence 
is genetical rather than cytological in nature. Although many cytologists 
have regarded the visible chromatin accumulations or granules as such units 
and have figured their division, others have found it increasingly difficult 
to see in these chromomeres anything significant in this conne ction. For the 
most part they seem to be quite inconstant in number and disappointingly 
irregular in behavior. 
In the somatic cells of Tradescantia, as in Vicia, chromosome splitting 
seems to be initiated by a series of axial vacuoles which quickly develop 
into openings through the homogeneous chromatin thread, and not by the 
division of chromatic granules supported by the thread. As the openings 
become more numerous the chromosome takes the form of a ladder, the 
rounds being represented by the material between the successive openings 
(fig. 2i). As the rounds or cross pieces gradually become thinner and 
finally break at the middle, thus completing the split, their material accumu- 
lates in two small lumps opposite each other in the two halves of the chromo- 
some. The ''paired granules" or ''divided granules" described by many 
workers are without doubt to be explained in this manner. Such being 
their mode of origin, it is difficult to assign to them the rank of morpho- 
logical units, or to attribute to them any specialized function in the heredi- 
tary process. In the microsporocytes of Vicia, however, the writer has 
observed chromatic granules of a much more definite character, but is not 
prepared to make any statement with regard to their significance. 
