ease: 
4 
1907) GATES—POLLEN DEVELOPMENT AND MUTATION 89 
these completely disappear, leaving no trace of the middle layers. The 
tapetum then lies directly against the epidermis. 
The presence of pollen grains in various stages of development 
when the tapetum persists, and their complete absence when the 
tapetum disappears early, shows that the failure of the tapetum to 
develop is at least one factor in the abortion of the pollen. Poxt’s 
conclusion that “die Zellvermehrung und relative Persistenz der 
Tapetumzellen hemmt die Pollenentwicklung” is evidently an error, 
for, in the great majority of cases at least, where the tapetum persists, 
it does not show abnormal radial elongation or growth to fill the locular 
cavity, the tapetal cells in older stages becoming instead very much 
flattened radially, as already seen. The cases where the tapetum 
does not persist, but where the cells of the middle layers grow in and 
obliterite the cavity, have already been described. In short, when 
the trpetum persists the middle layers break down, as is the usual 
condition in normil pollen development; but when the tapetum 
degenerates at an abnormilly early stage, the cells of the middle 
layers grow in and fill the loculus. 
History of the pollen mother cells 
PROPHASE AND SYNAPSIS 
The description thus far has dealt chiefly with the general mor- 
phology of pollen development in O. /ata; and the development of the 
pollen mother cells will now be taken up in greater detail. It isa 
Noteworthy fact that the mother cells of a given loculus show but 
little variation in their stage of development. Generally, however, 
there is a slight progression from top to bottom of a loculus; e. 8-5 
when the pollen mother cells at the top of a loculus are in pre-synaptic 
Stages, those at the bottom may be in the late post-synaptic stages 
of the prophase; or when the mother cells at the top of a loculus are in 
metaphase of the heterotypic division, those at the bottom may have 
reached the telophase with reconstructed nuclei. In comparing differ- 
ent loculi of an anther or different anthers of the same flower, the 
Variition is often much greater. Thus in some loculi the pollen 
Mother cells may be in synapsis; while in others of the same or another 
anther in the same flower, they may be in metaphase of the heterotypic 
