382 PROCEEDINGS OF SOCIETIES. 



especially of Strasburger" and Warming,! and also of Vesque, | 

 have yielded results sufficiently at variance with the older ones to 

 warrant a serious reconsideration of the whole question of 

 homologies. The main points may be put somewhat as follows: — 

 The embryo- sac is not the result of simple enlargement of one 

 cell ; a cell of an axial series of the ovule enlarges and cuts off two 

 or more cells from its apex — the remainder becomes the embryo- 

 sac, § and causes dissolution of the others as it enlarges, their 

 remains persisting as refractive ca^Ds for some time. The proto- 

 plasmic contents of the embryo- sac separates into two chief 

 masses, which pass to the opposite ends of the enlarging sac, a 

 large vacuole forming between. Each mass then suffers division 

 into four by planes cutting one another at right angles. In this 

 way eight nucleated masses of j)i"otoplasm without cell-walls 

 arise, three of which remain at each end of the embryo- sac, while 

 one from each end wanders towards the centre of the sac and 

 then fuses with its neighbour to form the nucleus of the sac. 

 Of the three anterior nucleated masses, two become elongated, 

 fit into the top of the embryo-sac as the " Gehiilfinnen," or 

 "synergidge" of Strasburger, and are probably what Schacht 

 described as the '-' filiform apparatus" ; their function is somewhat 

 obscure, but appears to be related to the act of fertilisation 

 between the end of the pollen -tube and the third nucleated mass, 

 which has rounded off as a large bright egg- cell or ovum, and is 

 suspended from the base and sides of the *' Gehiilfinnen." The 

 three posterior masses may not become completely isolated, or 

 they may remain passive, or some may disappear, or they may 

 multiply by division. Where endosperm is formed they appear to 

 enter into its formation. The author finds in Gyninadenia^W 

 Btitomus, Alisrna, Antliericum, and Rammcidus ovules confirmation 

 of these views. With reference to Warming and Vesque's view 

 that the origin of the embryo- sac is comparable to that of the 

 mother-cell of pollen, and that the eight nucleated masses 

 arranged in fours are therefore homologues of pollen grains 

 (i.e., sj)ores), the following facts are of imx^ortance, as failing to 

 support the theory: — 1. The diffluent cell walls which form 

 between the cells cut off by the embryo-sac mother-cell, are in no 

 way important on account of their resemblance to the diffluent 

 walls in pollen formation ; they also resemble the difiluent walls 

 formed in the suspensor (pro-embryo), or any other organ ra^oidly 

 absorbed, &c. 2. The division into fours by planes crossing at 

 right angles is common to many other processes besides pollen 

 formation, >.r/., the embryo itself. 3. Vesque's view as to the 



* ' Uebcn' Befrucbtung u. Zelltheilung,' and ' Die Angiospermen u. Gymno- 

 spermeii.' 



t ' Ann. des Se. Nat. Bot.,' 1878. 



+ ' Ann. des Sc. Nat. Bot.,' 18T8. 



\ Vesque, however, thought a fusion of at least two superposed cells produced 

 the embryo-sac. 



II Strasburger has partly worked this, so far as the first division of the embryo- 

 sac mother-cell are concerned. He also gives figures (>l\inthericum. Vesque's 

 account of! But omus is not supported by the author's drawings. 



