

396 BOTANICAL GAZETTE Ldecember 



JL verticillata they are almost spherical The determination of 

 their nature has been left to Mr. Gager, who announces work in 

 progress on them. I have found bodies in yl. tuberosa in the 

 latex tubes resembling those in the pollen of that species in 

 staining properties and in form, thus adding weight to the food 

 theory. In some pollen grains there are single spherical bodies 

 of varying size, resembling those in the groups in their staining 

 capacity {fig. 25). 



Concerning the pistil Schacht (10) states that the two carpels 

 roll in adaxially, each forming a loculus, while the fusion of their 

 tips results in the head (Corry's *' style table"). He also figures 

 an ovule showing a nucellus. Corry (5) has gone at length into 

 the formation of the head and especially into the origin of the 

 corpusculum and caudicules. Schumann, however, doubts his 

 conclusions. In regard to the development of the ovules almost 

 nothing has been done except in Cynanchum, in which Chauveaud 

 (3) reports that there is no nucellus, no integument, and only 

 one megaspore. 



The arrangement of the carpels in relation to the outermost- 

 sepal seems not to be a constant one, although usually one is 

 opposite. The carpels remain separate except at their tips. The 

 loculus formed by the rolling in of the edges opens in the head 

 a short distance above the stigma. The cleft on the inner sur- 

 face is never completely lost. 



Quite early, before ovules appear, before even the edges of 

 the carpels have rolled in to any considerable extent, the tips of 

 the carpels come in contact and begin fusion, forming the head. 

 This is about the time of the first division of the archesporium 

 in the anthers. The ends of the carpels enlarge enormously, 

 forming a somewhat spherical mass (the head) with a funnel- 

 shaped depression at the top, which is sometimes deep, some- 

 times shallow {fig. 28'), The fibrovascular bundles of the carpels 

 flatten out, and by the juxtaposition of their edges form a hollow 

 cylinder in the center. At the apex the cylinder spreads, forming 

 a funnel. Corry (5) figures it doubling back towards the base, 

 but diligent search revealed no such case. The stigma has long 

 been known to be situated just beneath the head, although some 



