Endospermbildung bei Angiospermen. 933 
morphologischen Wert und auf dieEndosp twicklung 
gehen. Die Spezialuntersuchung, die Sharp bei der Gattung 
Physostegia vorgenommen hat, wäre wohl geeignet, Licht in 
diese Sache zu bringen, dürfte jedoch in einem Punkt eine 
‚falsche Deutung einschließen. Der Embryosack von Physo- 
stegia entwickelt hier noch : vor dem jungen achtkernigen 
Stadium von der Antipodenregion aus einen lateralen Zweig, 
der schnell in das Integumentgewebe eindringt und, da er 
später fast das ganze Endosperm führen soll, den Endo- 
spermlobus vorstellt. In jungen Stadien schreibt Sharp ihm 
haustorielle Funktion zu. Für die Zeit der ersten Endosperm- 
teilungen gibt der Autor folgende Schilderung der Vorgänge 
im chalazalen Embryosackteil: »At about this time the anti- 
podal cell which lies nearest the sac cavity takes on an 
appearance quite different from the others. It becomes binucle- 
‚ate, the cytoplasm changes in character, staining more 
‚deeply and rapid enlargement causes its wall to become strongly 
eonvex. This enlargement continues until the cell bulges out 
Conspicousiy into the embryosac cavity and its wall thus 
Partitions off the small pocket in which it.lies with the 
other antipodals. In stages somewhat later it bears much 
resemblance to the first few cells of the endosperm, but the 
Possibility that it also is of endospermous origin is precluded 
by the fact, that it has been observed side by side with 
es undoubted endosperm nucleus resulting from the triple 
fusion.« »The function of the cell in question is in all proba- 
bility haustorial, recalling the behaviour of the basal antipodal 
m several genera of the Galieae.« An einer anderen Stelle 
Sagt er weiter: »The formation of the 'endosperm is of consi- 
derable interest, It is initiated by the division of the endo- 
‘perm nucleus, which occürs in the "narrow region of the 
ei near the haustorial cell.«...... The division is accom- 
Kar by a longitudinal wall running through the middle of 
ce . ».. Extension continues until it comes into contact 
als ia sac wall at or near the end of the endosperm lobe, 
i the micropylar lobe it was not observed to do SO, 
er 0 eb ends free.« »The cessation of endosperm Tortma- 
at an indefinite point results in nuclei being left free ın 
