392 BOTANICAL GAZETTE [December 



glands are almond-shaped, the flattening being due probably to 

 pressure, since those on the end of the common peduncle are 

 cylindrical. The base is more or less narrowed into a stalk ; 

 and those in the axils of the bracts are two or three times as 

 long, proportionally narrower, and more distinctly stalked. 

 Sometimes the epidermis at the tip is not glandular. Some of 

 those on the end of the common peduncle are as much as iioo^t 

 long including the stalk. They remind one here of multicellular 

 glandular hairs. The evidence as to the function of these 

 glands is negative. Their distribution and early appearance pre- 

 cludes the idea of any connection with cross pollination. I have 

 never seen an insect pay any attention to them, although I have 

 often watched for it. 



Chauveaud (3) reports that in Cynanchum the corolla, after 

 some elongation, unites with the stamens, but later becomes free. 

 The petals of the species studied are not at anytime united with 

 the stamens, although a transverse cut through a bud at an early 

 stage gives such an appearance, because the organs are closely 

 appressed and cut at an angle. At base the corolla forms a short 

 tube, which has probably arisen as a ring from the receptacle. 

 The dorsal surface bears stomata and becomes hairy ; and the 

 ventral surface just before the bud opens becomes papillose. 



The stamens have been much studied. Schacht (10) gives a 

 figure showing the beginning of hood and horn in A. Cormiti, 

 Corry (4) traces the detelopment of the wings in the same spe- 

 cies. He also follows the development of the sporangium, 

 wrongly deriving it from a single sporogenous cell. Chauveaud 

 (3) traces the development of the stamen of Cynanchum with 

 its hood or horn, and describes some features in the development 

 of the pollen. Strasburger (14), Gager (8), and the writer (7) 

 have given the history of the microspores. Stephens (13). how- 

 ever, seems to have seen the tetrad divisions ; and from an unpub- 

 lished paper now in my hands it seems that Miss Langley, now 

 deceased, of the University of Michigan, traced correctly the 

 pollen formation in 1S97. 



The stamens and petals early arise from a common ring 

 slightly elevated above the insertion of the other sets. From 



