﻿Eohrbach first of all divides the genus into two subgenera 

 (1) Behen, in which the leading character is "aestivatio petalorum 

 imbricativa"; and (2) Silene proper, in which the leading character 

 is "ffistivatio petalorum alternatim contorta." In connection with 

 this I should like to refer to an interesting letter from Alexander 

 Braun to J. Gay, preserved in Herb. Kew. In this letter diagrams 

 are given of the mode of overlapping of the petals in aestivation in 

 Silene Pumilio ; and quoting from a letter from Pacher, he says that 

 usually the petals are imbricate in activation in this species, but 

 that this is not constant, and occasionally they are contorted. 

 I consulted the late Sereno Watson (who has studied the North 

 American species of SUene) as to his views on the importance of 

 these primary characters. In a letter received from him only a few 

 weeks before his lamented death, he says; — "I have never con- 

 sidered the characters that you mention of any special importance, 

 and have paid them no attention. If 8. Douglasii and S. nivea are 

 to be separated from our other species on the imbrication of the 

 petals, it is evident that the division is not a natural one." Again, 

 "sestivatio alternatim contorta" is somewhat obscure, but Mr. 

 Watson points out that it appears to be illustrated in Eichler's 

 Bluthendiagramme, on p. 106, fig. 41a; and is what is called 

 "antidromy" or "heterodromy," the petals in the flowers of the 

 axis overlapping in one direction, and those of the branches in the 

 opposite. Now the species of Silene are difficult to discriminate in 

 living specimens, especially before the ripening of the capsules, and 

 still more so in the case of herbarium specimens; and were the 

 mode of imbrication of the petals, if such variation is to be depended 

 upon, to be insisted on as a primary character, it would of necessity 

 impair, if not stultify, the value of the result obtained from an 

 attempted examination of the greater part of the material afforded 

 by collectors and distributors. I propose, therefore, by slightly 

 modifying Eohrbach's arrangement, to base the primary sub- 

 divisions of the genus on the structure of the calyx ; and to 

 exclude from his subgenus Behen those few species in which the 

 two characters of a calyx with anastomosing veins and inflated at 

 the time of flowering are not associated, viz., S. pygmcea, 8. turgida, 

 and 8. Douglasii. I propose also to follow Engler & Prantl* in 

 considering each division of Silene proper as a subgenus, thus 

 making three subgenera instead of two. I would further suggest 

 GastrotUene as a substitute for Behen, which is somewhat of a 

 misnomer. Cucnbalus Behen L. and Behen vulgaris Monch, are 

 certainly superseded names for SUene infiata, the best known 

 species of the subgenus, but Silene Behen L. is a good species, and 

 belongs to another subgenus, so that the name as taken up by 

 Rohrbach is misleading. Gastrosilene is pseudhomonymic also with 

 Gastrolychnit, one of the sections of the genus Lychnis proposed by 



The limits of the genus will be more conveniently circumscribed 

 if Rohrbach's generic character of a unilocular capsule septate at the 



* Lie NUturlichen Pjtanzen/amilien, Theil iii. (1889), p. 70. 



