Igo] SHATTUCK—HETEROSPORY IN MARSILIA 23 
forming four spores (rarely 3 or 5), and a very large percentage of 
these developing, which are very constant in size, number, and 
behavior. 
After reviewing these facts we must conclude that the megaspore 
condition is a derived condition, that is, derived from the original 
homosporous condition of the ferns; that it is therefore more recent, 
and consequently more plastic, and more likely to yield interesting 
variations when subjected to experimental methods.t With these 
general facts in mind the experimental work was begun. 
Methods 
In attacking this problem it was first necessary to determine defi- 
nitely at what stage in its development the contents of the sporangium 
give positive evidence as to whether a megaspore or microspores are 
to be formed. Jounson (15) determined that the megasporangium 
is the first sporogenous tissue to be differentiated, but did not trace 
the development much farther. This has been done, and we find that 
all sporangia have identically the same development until the tetrads 
are formed (fig. 4), at which stage a very slight difference is observed. 
In the older sporangia (fig. 2), which are also the most centrally 
located, the four young spores of each tetrad show a marked tendency 
to hang together by strong protoplasmic strands, as figured by StRas- 
BURGER (26) and shown in figs. 5-10. These strands are the first 
recognizable morphological feature by which one can determine that 
@ Sporangium is to form a megaspore rather than microspores. They 
persist until the megaspore is quite mature, and in very many instances 
can be seen on the papilla of the germinating megaspore, which still 
subtends the three abortive members of the tetrad (fig. r0,a). This 
condition was figured by Wittramson and Scort (28) for C. Binney- 
ana, although they did not attempt an explanation. In the case of 
sporangia which were to form microspores it was observed that the 
protoplasmic strands are not so strong, and that the young spores, 
while held together during their early development, are in the end 
Separated completely from each other. The fact that the megaspores 
are held together by stronger protoplasmic strands was first noticed 
by Sacus in 1866 in Pilularia globulijera (22). 
' After this work was well under way Miss PFEIFFER (20) reported cases in Azolla 
where two megaspores have matured instead of one, which is the normal number 
