

534 ME. H. MARSHALL TVABD ON 



appears to be the wall of the cell at the base of the nucellus of fig. 6 ; 

 and the two nuclei and certain longitudinal fine lines in fig. 12 

 make one inclined to believe that two cells still exist side by side 

 {cf. PL XXIII. fig. 1) ; but the longitudinal wall between them is 

 being absorbed. 



In PL XXIV. fig. 2, we have the embryo-sac completed, and 

 its relation to the whole ovule shown ; while in PL XXIV. fig. 1 

 is a magnified view of the same sac, where an egg-apparatus is 



evident at the swollen (micropyle) end, while several dividing (?) 

 nuclei in the sac appear to have some relation to endosperm for- 

 mation. But I admit my inability to explain this figure and 



PL XXIV. fig. 3 in detail. 



To sum up. We have the embryo-sac in Pyrethrwm resulting 

 from the direct outgrowth of one cell, which sooner or later ob- 

 tains the upper hand in the struggle between it and its two or 

 three competitors in the nucellus, at the same time the epidermis 

 becoming used up by the developing sac. The process reminds 

 one, indeed, of what occurs in vascular Cryptogams, where one 

 macrospore, or spore-mother-cell, survives at the expense of its 

 neighbours. The details as to the method of forming the com- 

 plex contents of this embryo-sac I cannot at present describe. 



It remains to add, in explanation of PL XXIV. fig. 1 &c. that 

 the cells of the integument now closed around the sac become 

 altered, so that those in the layer next the sac take on the appear- 

 ance of cuboidal cells, and resemble an epidermis, so to speak, 

 while others grow very large, radiating more or less towards the 

 periphery, which consists of less-altered elongated cells. These 

 inner layers also have thickened walls (fig. 1); and this fact ? 

 coupled with their large size and radiate arrangement, explains 

 the peculiar appearance in PI. XXIV. fig. 2. 



Embryo-sac Sfc. of Anthemis tinctoria. (Plate XXIV.) 



I have selected one more Composite as an example of still 

 further reduction of the nucellus of the ovule. The figures will 

 explain themselves with the aid of a few remarks ; but it may be 

 premised that this short series is not complete ; for it is an open 

 question whether, after growing forwards, to fill up the space left 

 by the integument, the large cell in PL XXIV. fig. 9, undergoes 

 division. I have failed to observe this, and believe it does not 

 occur. 



