﻿11 



formulated and codified their rules, requiring here and there a 

 change of name. So that in a revision of the genus much has to 

 be reconsidered. 



In a previous paper* I essayed to point out that the 

 more the g garded in a sense restrictive of its 



Linnean definition, the better understood will be the affinities of 

 allied genera, and that a redistribution of many of the species 

 usually included in Lychnis will greatly facilitate the study of the 

 relations of the Silene group of genera. The genus Silene, as 

 understood and circumscribed by Alexander Braun, included a large 

 number of species which formed a very natural group sufficiently 

 marked off from other genera, as defined at that time, but very 

 difficult to form into subsidiary groups, on account of the absence 

 of well-defined primary and secondary characters within the limits 

 of the genus which might be utilized for the purpose. 



The point which I especially wish to draw attention to is the 

 extreme inconvenience of regarding the mode of praefloration in the 

 petals as a primary character. Boissier points out that Alexander 

 Braun first drew attention to this character in the genus, and the 

 speciousness of such a line of cleavage is best explained in his own 

 words, which are here translated from Fl. Orientalis, vol. i. :— "A 

 genus very difficult to break up into groups of species, since the 

 characters for defining sections are either absent or not strongly 

 marked. Thus Godron demonstrated that all forms of inflorescence 

 which were met with in the genus, the dichotomy, the panicle, the 

 unilateral raceme, are only modifications of the cyme, and that they 

 pass one into another in allied species, and even in plants of the 

 same species. The number of the nerves of the calyx varies in 

 plants otherwise alike in all their characters. Alexander Braun 

 pointed out a character depending on the imbricative (quincuncial) 

 aestivation of certain species, but this mode of overlapping of the 

 petals in aestivation in the genus is not really quincuncial (arising, 

 as it were, from spiral insertion of the petals), but is a deformation 

 of convolute estivation (arising from verticillary insertion), to which 

 in allied species and often in the same species it returns. The seeds 

 which are generally canaliculate on the dorsal surface may become 

 plane, then convex, and finally, in llcliusj; n,»t, carmate (f with the 

 rows of tubercles transformed into crested spines). With these 

 considerations, I have not proposed sections such as it would be 

 incumbent on a future monographer of the whole genus to specify, 

 but I have attempted to arrange the Eastern species in natural 

 groups, as far as I was able, according to general characters and 

 habit in the absence of definite and well-marked characters." 



Dr. Christ draws a harrowing picture of Boissier racking his 

 brains in his varied and repeated endeavours to select such cha- 

 racters as might form the basis for the main groups in the proposed 

 arrangement of the Eastern species ; how he tried first one character, 

 and then another, and afterwards a combination of two or more, 

 only to find that his hypothetical sections invariably either over- 

 lapped one another or failed to include some of the species. 



• Journ. Bot. 1893, p. 167. t Not in the original. 



