THE OVULE OF IMPATIENS 199 



Longo describes an abnormal development of the nucellus in 

 /. amphorata and in /. roylei, in which the micropyle becomes 

 filled during the development of the embryo, with a prolongation 

 of the nucellus. In /. amphorata this prolongation extends be- 

 yond the mycropyle, and in /. royeli, it forms a large branching 

 haustorium, penetrating deep into the tissue of the funiculus. 

 In the haustorium the suspensor is imbedded. The outline draw- 

 ings which accompanied Longo's paper did not make the nature 

 of the haustorium clear in support of his statement, but for a 

 number of reasons, I believe that they are not composed of nucel- 

 lus cells. In the first place, there are no nucellus cells remaining 

 in the micropylar end of the ovule in some of his drawings, and 

 only a few fragments in others, at the time that the haustoria are 

 developing. When the nucellus elongates, it is naturally during 

 its active period, as in Euphorbia corollata^ and not when it is 

 disintegrating. Now the tissue in the drawing, in so far as its 

 nature is shown, looks more like endosperm than like nucellar 

 tissue, and it seemed that that was more likely its origin since 

 there are many similar cases, e.g., Globularia cordifolia, Stylidium 

 squamellosum, Biblis gigantia, Capsella bursa-pastoris, and others. 

 After reading Longo's paper, I made some slides of more advanced 

 stages in my material and found a similar haustorium, though it 

 was a short one, extending through from one-third to one-half 

 the length of the micropyle. The tissue of this haustorium is 

 clearly endosperm because (1) The nuclei are like the endosperm 

 nuclei; (2) the staining is identical with that of the endosperm, 

 and different from that of the nucellus; (3) the tissue is con- 

 tinuous with the endosperm, the drawings make the identity 

 of the tissue clear (fig. 2, Q and S). There seems to be no ques- 

 tion that the tissue in the micropyle of /. amphorata and of /. 

 roylei is also endosperm. 



There is no parietal tissue formed, a common condition among 

 Archichlamydeae and universal among the Sympetalae.^ Four 

 megaspores are formed, the chalazal one developing at the ex- 

 pense of the other three (fig. 1, J). The nucleus divides as the 



^ Coulter and Chamberlain, I. c. 

 * Coulter and Chamberlain, I. c. 



