
388 BOTANICAL GAZETTE [NOVEMBER » 
1904 OVERTON (18) investigated Thalictrum, finding the pollen 
normal. In some instances there was a reduction division in the 
development of megaspores, while others developed without this, a 
case exactly parallel to that of Marsilia as shown by STRASBURGER 
(24) in 1907. In Taraxacum officinale Jurt (12, 13) in 1904 and 
1905 states that the embryo sac mother cell divides only once, the 
inner daughter cell developing the sac, but no reduction division 
takes place. The reduction division does take place in the micro- 
spore development. 
ROSENBERG (22) in 1907 found in Hieracium that the reduction 
division takes place in the microspore development, but in some 
species, at least, they are disintegrating even at this time, and 
the mature anthers are empty. In some species the megaspore 
development is occasionally normal, giving a haploid gametophyte; 
but it usually happens that some sporophyte cell (epidermal, 
chalazal, or nucellar cell) produces the embryo sac. So this is a 
- case of apospory in the development of the sac; and as a diploid 
egg is organized, it is also a case of apogamy in the development of 
the embryo. The work of WINKLER (28) in 1904 and of STRAS- _ 
BURGER (26) in 1909 with Wikstroemia shows reduction in the micro- 
spore development with but little pollen maturing, while reduction 
does not take place in the megaspore development. In 1910 
STRASBURGER (27) added Elatostema to the apogamous angiosperms. 
It also has the reduction in the pollen development with much 
irregularity in the pollen, some anthers producing pollen of normal 
appearance. The mother cell of the ovule begins the reduction 
division but returns to the vegetative mitosis; but in some as 
reduction does take place. From these it seems that reduction 
always takes place in the microspore development, the failure being 
only in the megaspore series, and in some species it is only partl 
here. 
It will be seen that normal pollen is reported only for T halictrum, 
some species of Taraxacum, and Atamosco, the others all having 
little pollen or defective pollen. The female gametophytes for 
four genera are diploid, while in the other three both haploid and 
diploid female gametophytes are developed; but it is not knoe 
that these haploid eggs develop parthenogenetically. According to 

