THE EMBItYO-SAC IN ANGIOSPEKMS. 537 



but, from analogy with Senecio, in which the processes are so far 

 very similar, it is probably from the lower one *. 



In concluding these remarks on the drawings accompanying 

 them, the following general statements must be taken for what 

 they are worth, as suggestions rather than any theory. They are 

 not based upon these drawings only, but, besides having reference 

 to the papers quoted, apply to results from Gymnadenia, Senecio, 

 Ranunculus, Myosurus, AntTiericum, Bisum, Agrinionia, and 

 others. 



Whether the nucellus of the ovule consists of few or many cells, 

 a definite relation can be traced between the mode of division of 

 the cells and the shape of the whole ovule. As the latter increases 

 its volume by growth along given lines, this is brought about by 

 division perpendicular to those lines ; and (at least for a long 

 time) the same applies to growth along the longitudinal axis : the 

 central series of cells are divided by transverse walls. 



In some cases (e. g. JRosa, Lupinus, Biitomus, CEnothera, &c.) 

 the nucellus consists of radiating groups of cells starting from a 

 common centre about the lower part of the whole mass, which, so 

 far as the method of division is concerned, behave similarly, and 

 might be considered of equal value. But one (Butomus &c.,)t, or 

 more (JRosa &c.) cells, usually at the base of its group of cells, 

 becomes distinguished at an early date, by the accumulation of 

 protoplasm &c. in its cavity, as the forerunner of the embryo-sac 



the " embryo-sac mother cell." In other nucelli these equiva- 

 lent groups of cells are fewer and more insignificant ; in such 

 cases as Anemone, for instance, one might imagine the basal cells 

 of the side groups in Butomus to have become absorbed into the 

 rudiment of the embryo -sac mother cell; while in Composite and 

 others we haye the nucleus reduced to the axial row only (Sene- 

 cio &c.) or to a group of three or four cells. In some cases (Senecio, 

 Lobelia, Verhascum, Orchids, &c.) this cell divides as the nucleus 

 elongates j in others (Pyrethrum, Antliemis) the cell or cells merely 

 elongate for the short distance allowed them. 



Whatever be the morphological significance of the mother cell 

 of the embryo-sac, and therefore of the embryo-sac itself, then, 

 we cannot derive it from the manner of division exhibited by its 



Cf. also Warming's account c 

 t In Butomus and Lupinus, ho* 

 other cells around the central one. 



this affecting 



