222 KirKWoop: POLLEN-FORMATION IN CUCURBITACEAE 
But few plants of the Czcurbitaceae have received attention 
from this point of view. Mirbel ® studied the development of the 
anther of Cucurbita Pepo and demonstrated the principal facts of 
the process. In small buds 2 mm. long he found no trace of the 
locules, but in a slightly later stage he was able to discern the 
spore-mother-cells and the tapetum. His figures of this condition 
represent the anther-wall, composed of four layers of cells under 
the epidermis including the tapetum. In buds 3 or 4mm. long 
an additional layer of cells was detected in the anther wall. The 
development of the pollen-mother-cells, the formation of the 
tetrads, and the differentiation of the pollen-grain are well de- 
scribed and figured. 
Naegeli described certain features in the formation of pollen 
in Cucurbita and in Bryonia dioica. He seems to have observed 
the first division of the microspore-nucleus in Cucurbita but not 
to have interpreted correctly what he saw. He discusses moreat — 
length the differentiation of the exine and the behavior of the 
intine upon the germination of the grain. 
To Warming,” however, weare indebted for an accurate study 
of the development of the anthers of Bryonia alba and Cyclanthera 
pedata. In both these cases the first periblem layer of the anther 
divides by periclinal walls. From the outer cells thus formed is 
developed the anther-wall by succeeding periclinal divisions, and 
the inner cells become the archesporium, In Bryonia the arche- 
sporium is a single layer of cells which later forms a mass of spore- 
mother-cells. Warming says that, as seen in transverse section, 
the pollen-mother-cells form one to several rows in each angle of 
the anther. Not all the cells cut off toward the inner side in the 
division of the hypodermal cell become mother-cells, and some- 
times those that do become mother-cells do not divide again until 
the formation of the tetrads. In Cyclanthera the inner cells result- 
ing from the division of the first periblem layer as a rule do not 
divide again but form mother-cells by growth. 
Thus the evidence indicates that in these plants the usual 
order of development of sporogenous tissue in seed plants is ad- 
hered to, but the subsequent history of the archesporium may 
vary, as it appears at present, in accordance with the form and 
structure of the anther, in some cases the original archesporial 
