424 The Ohio Naturalist. [Vol. VI, No. 3, 



were Anilin Safranin and Gentian Violet, Heidenhain's Iron- 

 Alum-Haematoxylin, and Delafield's Haematoxvlin, the latter 

 perhaps giving the best results. Care had to be taken with it 

 and the Tron-Alum-Hacmatoxylin as the embrvo sacs and em- 

 bryos stained so deeply that it was difficult to make out the 

 details unless a large part of the stain was removed. The stages 

 just after fertilization were quite difficult to observe as the pollen 

 tube discharges a quantity of material which stains very deeplv 

 and obscures the embryo sac structures. 



Orientation for sectioning was not difficult as the ovularv 

 when cut crosswise gives longitudinal sections of a number of 

 ovules. For the older stages only a portion of the ovularv 

 could be sectioned on account of its size. 



The cross section of the very young ovulary shows the 

 placentae with minute protuberances which represent the incip- 

 ient ovules (Fig. 1). The carpel has three placentae, and the 

 ovtiles are delveoped in six rows which are usually double, but 

 this is somewhat irregular. The tip of the ovule remains 

 straight for only a short time after the appearance of the arche- 

 sporial cell (Fig. 6). The cells along the outer margin begin to 

 divide more rapidly than those of the inner side. This unequal 

 growth causes the ovule to turn, and this process continues until 

 the micropyle is brought close to the funiculus. Before the 

 megasporocyte has divided and before the integuments have 

 grown over the nucellus the ovule has curved half the distance, 

 and the normal anatropous condition is practically attained 

 w^hen the ovule has reached the megaspore stage (Fig. 5). At 

 this time the characteristic beak which develops at the tip of 

 the nucellus is already becoming prominent. 



The integuments lengthen greatly forming a long narrow 

 micropyle into which the neck-like process of the flask-shaped 

 nucellus projects, even to the tip of the integuments. 



The archesporium is as usual a single hypodermal cell that 

 terminates the axial row of the nucellus. It can easily be dis- 

 tinguished from the surrounding cells by its greater size and_ 

 deeper color due to the denser protoplasmic contents (Fig. 6).' 



By a transverse division the archesporial cell gives rise to two 

 cells, the megasporocyte and primary parietal cell (Fig. 7). The 

 latter continues to divide by both periclinal and anticlinal walls 

 thus forming the parietal layer (Fig. S) which remains persistent 

 and with adjoining cells keeps on dividing to form the long beak 

 of the nucellus (Fig. 22). 



The megasporocyte is carried down into the tissue quite a 

 distance by the development of the parietal layer before any 

 division occurs. The division of the megasporocyte is normal, 

 giving rise to four equal megaspores (Fig. 10). The potential 

 megaspores soon begin to dissolve and the lower or functional 



