PHANEROGAM1A : FLOAVERS. 



401 



here a part of the integument (of the micropyle); in Lathraa, which 

 plant has altogether a strangely aberrant form of seed-bud, there is 

 formed, not merely here, but at the opposite extremity, a caecal, sac like 

 appendix (fig. 238.). 



Lastly, in the Santalacece it emerges from and lies quite free outside 

 the micropyle as a sac of varying length.* 



The formation of several embryo-sacs in Viscum, already mentioned, 

 is at present peculiar to it. Meyen has the merit of first directing at- 

 tention to it, and almost completely unfolding the particulars. In the 

 beginning the cellular tissue of the pedicel of Viscum is perfectly homo- 

 geneous ; by degrees an abundance of fluid is secreted between the cells 

 lying in the axis ; they separate from their union, and a kind of cavity 

 is formed in which they lie loose. This I formerlyf, before I had suf- 

 ficient material for the history of development, erroneously stated to be 

 the embryo-sac. Two or three of the loose cells in this cavity then ex- 

 pand into the form of tubes, which gradually displace the others (fig. 222.). 

 All the rest, the gradual formation of cellular tissue, and the subsequent 

 processes in the formation of the embryo, agree perfectly with those of 

 other plants. \ Only in a few cases does the embryo-sac, even at this 

 period, become wholly filled with cellular tissue, but in a great many 

 plants there begins at this time (as always happens later) a process of 

 cell -formation, which constantly commences at the circumference of the 

 cavity, and proceeds inwards ; when a few layers of cellular tissue have 

 been formed in this way upon the periphery, they represent what Mirbel 

 has called the quartine, and from not having completely traced their de- 

 velopment, has described as a fourth integument formed between the 

 nuclear membrane and the embryo-sac. It is not a rare occurrence, and 

 in the very family of Cruciferce cited by Mirbel is easily traced in the 

 way I have described it. There can be no question of integuments here. 

 I will merely remark in passing that every embryo-sac subsequently be- 

 comes gradually filled with cellular tissue, which either becomes wholly 

 displaced by the advance of the embryo or remains as endosperm (al- 

 bumen); whether it appears some- 

 what earlier or later makes no 

 difference. In the Conifer ce cellu- 

 lar tissue is likewise formed in the 

 embryo-sac, which, however, be- 

 comes so arranged that from three 

 to six larger cells immediately 

 beneath the part turned toward 

 the nuclear papilla, and clothed 

 only by the embryo-sac on the out- 

 side, become especially strongly 

 developed. The layer of cellular 

 tissue which bounds these cells 

 acquires an epithelial aspect, so 

 that these cells appear like defi- 

 nitely bounded little sacs (Robert 

 Brown's corpusctifa.) I have 



* Griffith on the Ovule of Santalum, Linn. Trans, vol. xviii. 



t Wiegra. Archiv, 1839, vol. i. p. 212.; Schleiden's Beit, zur Botanik, vol. i. p. 21. 

 pi. 2. fig. 15. 



| Link (Wiegm. Archiv, 1841, vol. i. p. 393.), in opposing DeCaisne's very valua- 



151 4 bits exceha. A, Seed-bud in longitudinal section ; />, nucleus ; r, embryo-sac ; 



D D 



251 



