164 PROCEEDINGS OF THE ACADEMY OF [1892. 



in the form I found, exposing the flowers, that led to the discovery. 

 Since then I have examined herbarium specimens from a very 

 wide range of territory, and find these hidden flowers always with 

 this species, and in many cases giving such a gland-like character 

 to the base of the ochrea that it is inconceivable how the production 

 of these flowers should have been overlooked so long. I gathered 

 some specimens in the streets of Washington, and showed them to 

 the members of the American Association in attendance there ; and 

 Mr. T. H. Kearney, Jr., of the University of Tennessee, has since 

 confirmed the discovery by a note in the November issue of the 

 Botanical Gazette. He found them in Knoxville, Tenn., as late as 

 Sept. 24th. His excellent drawing, accompanying his note, shows the 

 cleistogene flowers are much closer down among the roots than I found 

 them. Mr. Kearney could find them in no other species, nor have I 

 been able to do so. A close examination of many in the large herbar- 

 ium of the Academy, gave no indications of this tendency in any 

 other species. As already noted there seems no fertile flowers in any 

 Polygonum, except from cleistogene flowers at the apex of the 

 branches. This species has them specially in the axils as well as in 

 the terminal spikes. 



In examining a large number of specimens in herbaria in connec- 

 tion with this question, it is interesting to note how often P. acre is 

 confounded with P. Ilydropiper. For all the differences noted in 

 our botanies, I think they are more closely related than supposed. 

 I once thought I could easily decide the differences by the seeds, 

 but I have not unfrequently found flattish seeds in P. acre, that 

 could not be distinguished from the usual flattish seeds of P. 

 Ilydropiper. In forms, however, that we would certainly refer to 

 the latter species, no tendency to the axillary cleistogamous flowers 

 can be detected. 



On the direction of growth in Cryptogamic Plants. 



No one has yet been able to present an explanation of the direc- 

 tion of growth in flowering plants, that will stand the test of 

 criticism. Growth lias a general upward tendency, though in the 

 same tree we have varying directions. INTany Conifene have one 

 perfectly vertical central stem or leader, while the side branches 

 may be wholly horizontal, or at various angles uniform in each 

 species. Occasionally individuals will vary from the normal line of 

 direction, and present angles wholly different from that jjrevailing 



