89 
give rise to the embryo sac, in other instances the micropylar 
one constantly gives rise to the embryosac (Solidago serotina 
(Patm 1914a), — In the few instances recorded where only one 
out of four “syncytial” megaspores develops into the embryosac, 
it is always the micropylar megaspore nucleus that gives rise 
to the nuclei of the embryosac, so fii. in Anthemis tinctoria 
(Homeren 1915) and Pyrethrum corymbosum (Patm 1915). 
In the second type known as the Scilla type, two megaspores 
join in the formation ot the gametophyte. Here a wall may be 
laid down in the heterotypic division, this is the rule in Hrigeron 
unalaschkensis, 2. Coulteri (Hotmertn 1919) and perhaps also in 
Larthenium argentatum (Kirxwoop 1910); it also occurs as an 
exception in Zunacetum vulgare (Paum 19140, 1915) or a wall is 
not formed as in Tanacetum vulgare (Patm 19146, 1915). The 
two basal megaspores give rise to the embryosac in Hrigeron 
unalaschkensis, in Tanacetum on the other hand the two micro- 
pylar ones give rise to the embryosac. In the latter plant the 
basal megaspores do not degenerate and may even occasionally 
divide once. 
To the Scilla type belongs also Vittadinia, as will be shown later. 
The third type comprises the cases where all four megaspores 
take part in the formation of the gametophyte; here no walls 
separate the megaspores. In this connection it is worth men- 
tioning that in all known instances of this kind of combined 
megaspore development in Compositae, the resulting embryosac 
follows the so called Peperomia type and consists of sixteen 
nuclei. Examples in the Compositae are Erigeron eriocephales, 
E. politus (Houmeren 1919) and Pyrethrum parthenifolium v. aureum 
(Pau 1914, 1915). According to Kirxwoop (1910) in some species 
of Parthenium an eight nucleate embryosac developes from four 
megaspores thus forming en embryosac of the Lilium type. 
Further investigations are, however, needed to confirm this point. 
In the course of an investigation upon some special features 
of the antipodal development of some Asterae, Vittadinia triloba 
D. C. was also investigated. The antipodals not being of any 
marked interest in the line of research undertaken, the early 
Ann. Jard, Bot. Buitenz, Vol. XXXII. 8 
