1915] MICHELL—STRIGA LUTEA 127 
embryo sac by their much smaller size. This case is yet another 
to be added to the ever-growing list of plants in which double 
fertilization is known to occur. 
Endosperm formation 
The endosperm is initiated by cell formation. Striga lutea thus 
conforms with the type largely represented in the Sympetalae and 
almost universal in parasites and saprophytes (5). 
The first division of the primary endosperm nucleus is immedi- 
ately followed by the formation of a transverse wall, dividing the 
- embyro sac into two chambers (fig. 19). The nucleus of the chalazal 
chamber divides once, but no wall is formed, and endosperm is 
never produced in this chamber. Its function is clearly haustorial, 
for soon after the division of the nucleus the chalazal end of the 
sac grows down into the integument, and finally curving upward 
reaches the outermost layer of the integument (fig. 22). Fig. 21 
shows a somewhat earlier stage in which the remains of the antip- 
odals are seen at the end of the haustorium. 
The nucleus of the micropylar chamber divides three or four 
times, each division being accompanied by a wall transverse to the 
long axis of the sac (fig. 20). After this, walls appear in various 
planes and the original transverse walls are soon obscured. No 
definite micropylar haustorium is formed, though the cells of the 
endosperm grow a short distance up the micropyle, surrounding 
the suspensor, and are probably to be considered as having a 
haustorial function (fig. 23). 
The endosperm cells around the suspensor (fig. 23) and those 
formed at the base of the original micropylar chamber (fig. 22) have 
dense protoplasm and stain far more deeply than the rest of the 
endosperm cells. This may be due to the fact that they are active 
in passing the food, which is gradually being absorbed from the 
integument by the haustoria, to the developing embryo. In these 
ovules there is no distinct tapetal layer round the embryo sac. 
In the endosperm formation and in the development of haus- 
toria there is a remarkably close resemblance between Siriga lutea 
and Linaria alpina. Scumip (13) in his account of Linaria alpina 
might well be describing Striga lutea, the chief difference being that 
