Notes on Recent Literature. 
384 
megaspores is formed, it is almost invariably the one at the 
chalazal end of the embryo-sac that develops, 1 while the others 
degenerate, and there is no adequate reason why this arrangement 
should he reversed and all four megaspores germinate together to 
form the embryo-sac in the many cases where this tetrad is 
suppressed. 
4. In Peperomia, cell-plates or evanescent cell-walls are 
formed between the nuclei in the first two divisions, indicating that 
they should be regarded as nuclei of once-walled spores. 1 Exactly 
similar, and very suggestive in this connection, is the case of 
Sniilaciua, where four walled megaspores are formed, arranged in a 
row or tetrahedrally, and then their division walls are absorbed, so 
that the megaspore-nuclei occupy a common cell-cavity. In 
Sniilacina stellata, each megaspore nucleus then divides once, and the 
resulting eight nuclei group themselves to form an embryo-sac of 
the usual type. In Sniilacina racemosa, the two lower megaspore- 
nuclei do not divide, but the two upper each divide twice, and the 
resulting eight nuclei similarly organise the embryo-sac. This fact 
strongly suggests that the first four nuclei formed in the embryo- 
sac, in other cases where no walled spores are formed, are them¬ 
selves spore-nuclei; and similarly that when the mother-cell has 
only divided once and the embryo-sac originates from one daughter¬ 
cell, the two nuclei formed by the first division within this embryo- 
sac (the homotype division) are the nuclei of two spores. 
On the other hand, cell-plates or walls are not reported for the 
reduction-divisions of the other sixteen-nucleate forms. Moreover, 
“ an evanescent cell-plate is frequently observed during the three 
free nuclear divisions by which the eight-nucleate stage of the 
embryo-sac is reached, 3 ” and the occurrence of such plates cannot 
be considered as a criterion of whether the nuclei between which 
they form are those of spores. 
In this brief review, many of the arguments and supporting 
details used by the upholders of both sides of the question have had 
to be omitted. Even so, enough has been said to show the difficulty 
of proving or disproving either view in the present state of our 
knowledge of the embryo-sac. On the whole, it may be said that 
the trend of the majority of recent papers is either to leave the 
question open, or to maintain Coulter’s theory. 3 This theory 
certainly appears to throw light on certain hitherto obscure 
deviations from the normal type of Angiosperm embryo-sac. Thus 
it may be held to elucidate the anomalous embryo-sacs described 
by Campbell in Aglaonema commutatuni 4 and NepJithytis Liberica 5 
(species of the Aracete). In these forms, the number of “ sporo- 
genous cells ” is variable ; each “ sporogenous cell ” is a potential 
embryo-sac, and usually more than one begin to develop. The 
number of nuclei in the mature embryo-sac maybe increased to “at 
least fifteen,” their polarity is imperfectly marked, multiple nuclear 
fusions are seen, and the usual structures of the embryo-sac are 
not clearly differentiated. It is significant to note that Campbell 
observes of one species “ it looked sometimes as if the structures 
present at the time of fertilization were the combined products of 
two or more of the primary embryo-sac's 3 ” —and again of the other 
1 Brown, 190S. 2 Coulter and Chamberlain, 1903. 
3 See however Campbell, 1909. 4 Campbell, 1903. 6 Campbell, 1905, 
c Ibid, p. 339. 
