426 The American Naturalist [May, 



transverse to the length of the tube, the two new nuclei are so placed 

 that. the lower one is preceded, and the upper one followed, by its 

 two spheres. Both of these nuclei are accompanied by their proto- 

 plasm and each has been seen to have its twelve chromatin rods. If 

 the vegetative nucleus has not already disappeared, it may be distin- 

 guished from the sexual elements by the difference in reaction to 

 methyl green and fuchsine. 



The steps by which the differentiation of the female factor arises 

 may be rapidly sketched as follows : A large cell in the axial row of 

 the nucellus is the mother-cell of the embryo-sac. In all the cell 

 divisions of the nucellus up to the moment when one of the cells 

 becomes conspicuously larger and ceases to undergo further division 

 the nuclei have all shown twenty-four chromatin rods. The embryo- 

 sac grows rapidly so as to come to occupy a large part of the nucellus, 

 sometimes descending touanN the center of the same. Its nucleus 

 contains one or more large nucleoli in the meshes of the tangled 

 ribbon made up of two series of chromatin granules. The two directive 

 spheres lie in contact with the nuclear membrane and the cytoplasm i 



now seen radiating witl regard to the position of those spheres. 



The ribbon now break* into rods ami the substance of the nucleoli so 

 changes as no longer to be distinguishable from that of the chromatin 

 segments, the number of which is now twelve not twenty-four. The 

 spheres become active, go to the poles and are surrounded by cyto- 

 plasmic rays. At first the rods are scattered in a disorderly manner 

 through the forming spindle, but are soon carried and properly 

 oriented by the achromatic threads about the equator so that there 

 may now be seen a rod to each large thread, there being however 

 some more delicate intermediate threads to which no rods are attached. 

 Thus is formed the spindle and disk. The rods are easily seen to con- 

 sist of longitudinal halves and in the further change they split at the 

 ends directed towards the spindle axis, one-half appearing to glide up 

 a thread towards one pole, the other half moving down towards the 

 opposite pole until they finally part and become new groups of twelve 

 contorted segment., shaping themselves into two new nuclei. The 

 centrosomes in the meantime have doubled, forming two spheres at 

 each end of the old nuclear axis. The new nuclei so formed appeal • 

 all respects quite similar and they move to opposite ends of the embryo- 

 sac. The lower one especially gains in size, and both continue to 

 divide. But in the next division the lower one is seen to have not 

 twelve, but sixteen or twenty rods, whilst the upper one and it- deriva- 

 tives still have but twelve. There are now four cells or nuclei above 



