1906] OVERTON—THECOTHEUS PELLETIERI 483 
spores of many of the pteridophytes and spermatophytes. The spore 
not only begins its germination while inside the sporangium, but while 
‘it is still iaside the mother cell. The ascus is not only to be looked 
upon as a mother cell, but also as a mother cell which functions 
directly asa sporangium. The prevalent impression is that the history 
of the gametophyte begins with the division of the mother cell and 
ends in the act of fertilization. The ordinary product of the division 
of the spore mother cell is four spores, or in typical Ascomycetes 
eight spores. In Lilium the first mitosis of the mother cell is hetero- 
typical, while the second corresponds exactly in all details to the 
second division where normal tetrads are to be formed. We have 
here a double division completed inside the mother cell, and con- 
sequent germination to form a gametophyte inside the mother cell. 
It does not seem inconsistent, therefore, to think of a mother cell 
containing a gametophyte, or that the reduction divisions may not 
give rise directly to morphological spores. These nuclei are game- 
tophytic in character and can give rise to gametophyte structures in 
the embryo sac. It is not absolutely essential, therefore, that the double 
division result in spore formation. JuEL (:00) found that in Carex 
acuta the usual double division occurs in pollen mother cells, com- 
plete cell plates being formed which are later resorbed, so that four 
nuclei lie iaside the wall of the mother cell, three of which disin- 
tegrate, the fourth forming a single functional microspore. In Fuchsia 
(WILLE ’86) as many as fourteen microspores have been reported 
from a single mother cell, while more or less than four have been 
found in several other forms. STRASBURGER and JvuEL have also 
counted numerous microspores formed from a single mother cell. 
It would appear, therefore, that the double division is necessary, 
but that the number of spores ultimately formed is very variable. If 
in Fuchsia the walls of the microspores were eliminated, a striking 
s resemblance to the sixteen-nucleate stage of Thecotheus would result. 
The number of spores formed and the time of their formation seems 
to be very variable, but this does not interfere with our conception of 
the alternation of generations in the flowering plants. That thirty-two 
free nuclei are formed in Thecotheus before spore delimitation occurs 
is therefore no more striking than that tetrads are not formed as a 
result of the double division in Lilium, or that more than four micro- 
