148 MINNESOTA BOTANICAL STUDIES. 



velopment the embryo is not always attached centrally at the 

 anterior end of the sac, nor does it often lie in an exactly longi- 

 tudinal direction in the sac. 



It has already been stated that the definite nucleus in Riimex 

 does not divide till after fecundation of the egg, an observation 

 based upon examination of two or three hundred sacs. It is 

 worthy of note that the endosperm nuclei observed while divid- 

 ing in any given sac were all in the same mitotic phase (Figs. 

 30 and 34). The last figure shows only one of four nuclei ob- 

 served dividing. The endosperm nuclei were not so numerous, 

 at the stages studied, in Rumex ve?'ticiUatus as in Rimiex salici- 

 foliiis (Fig. 33) ; nor were they yet enclosed in cell walls in 

 either plant. 



Relation of the Ganietophyte to the Macros^orangiinn. — I 

 have studied carefully the position of the base of the sporogen- 

 ous tissues and derived embryo sac with reference to lines con- 

 necting the points of origin of the seed coats in order to as- 

 certain how much of the enlargement of these structures is 

 associated with a downward growth and consequent crowding 

 of the tissues of the macrosporangium and how much is accom- 

 plished by upward growth, keeping pace with the growth of 

 the nucellus. The position of the base of sporogenous tissue, 

 and later of the sac, with reference to these lines is not always 

 quite the same at any particular stage of development ; but by 

 comparative study, safe conclusions have been secured. Be- 

 tween the base of the archesporium and the lines connecting 

 the supposed points of origin of the future seed coat are five or 

 six cells. After the tapetum is cut off (Fig. 2), there are only 

 three or four cells between the base of the mother cell of the 

 macrospore and the lines. By the time of division of the tape- 

 tum (Fig. 3) the base of the mother cell is within two or three 

 cells of these lines. During this time the nucellus has increased 

 in length very little, its increase in size being principally in 

 width. Consequently, this fact, together with the relative posi- 

 tion of the base of sporogenous tissue and the lines at various 

 stages of development, indicates that the sporogenous tissue has 

 grown downward in the nucellus. As no evidence of absorp- 

 tion of cells was seen at this time, I conclude that this down- 

 ward growth is accomplished by crowding downward and out- 

 ward the subjacent layers of cells of the macrosporangium. 

 However, the effects of the crowding were so distributed among 



