McAllister—Cytology and Embryology. 625 
plain that the embryo-sac of Smilacina racemosa is formed from 
two individual macrospores and the method of the formation of 
the macrospores here suggests strongly that all those embryo- 
sacs which are formed from one of the daughter cells formed by 
the first division of the emryo-sac mother cell are to be regarded 
morphologically as formed from two macrospores. 
2. The development of the microspores. 
In order to more accurately interpret the figures found in the 
reduction divisions I have first taken up a study of division in 
the somatic cells from which the spore mother cells are formed. 
The resting nuclei of the cells just preceding the last premeio- 
tic division are similar to those in the other embryonic areas of 
the sporophyte. The characteristic reticulum of small chroma¬ 
tin bodies connected by linin strands is clearly seen (Fig. 1). 
With the initiation of the prophases of the last premeiotic 
division, the chromatin bodies lose some of their tendency to 
take the blue stain and take the red stain more readily. At this 
time they become arranged in more or less definite rows, con¬ 
nected by the linin elements, which take less stain than the chro¬ 
matin bodies. In this characteristic manner the resting reticu¬ 
lum becomes transformed into a kinky, closely packed spirem. 
Whether there is ever a union of two or more net knots of the 
resting reticulum to form a single chromomere could not be defi¬ 
nitely determined,—but from the irregularity in size of the 
chromatin masses of the reticulum and their comparative regu¬ 
larity in diameter in the spirem it seems very probable that 
some rearrangemexit is necessary to form the uniform chromo- 
meres. 
I was unable to find a paired condition of like chromosomes 
in the division figures of sporophytic nuclei such as Strasburger 
and others have seen in plant cells recently investigated by 
them. Their figures representing the pairing of somatic chro¬ 
mosomes are by no means convincing and it would seem that 
the foreshortening of those chromosomes seen from the end as 
well as the possible unequal rates of contraction of the chromo¬ 
somes in the stages following the transverse segmentation of the 
