278 ME. A. W. BE3TCETT ON CLEISTOGAMIC ELOWERS. 



I. glandulifera, Am. (Ceylon), does not appear to bear cleis- 

 togamic flowers. I observe that, in his ' Flora of British India,' 

 Sir J. D. Hooker states that the capsule is unknown. 



General Remarks. 



It will be observed that cleistogamic flowers are of two kinds : 

 first, those which hardly differ from the perfect open flowers in 

 any other respect than the partial or entire suppression of the 

 corolla and the closing of the calyx ; secondly, those in which 

 there is a distinct modification of one or more parts of the flower 

 to aid in the process of self-fertilization. The first kind, which 

 . may perhaps be called " homocleistogamic," are well illustrated 

 in the case of Oxalis AcetoseUa and in that ofDroserarotundifoha, 

 as described by Mr. Darwin ; the second kind, or " heterocleisto- 

 gamic," by the dimorphic species of Viola and Impatiens. There 

 is not of course any exact line of demarcation between the two 

 kinds, and intermediate forms occur, as in Viola sylvatica ; while 

 the homocleistogamic pass insensibly into the" normal open flowers. 

 I was at first disposed to think that these two kinds of closed 

 flowers might have arisen in different ways, the one by degrada- 

 tion, the other as a rudimentary form of the organ. But that the 

 heterocleistogamic flowers are really degraded normal flowers is, 

 I think, sufficiently shown by the frequent occurrence in them ot 

 points of structure which are absolutely useless to them when 

 self-fertilized, and which can only be regarded as survivals from 

 the normal flowers — as, for instance, the abortive stamens m 

 Viola ; the occurrence of the short outer whorl of stamens m 

 Oxalis ; and, as Mr. Darwin points out, the fact that in the cleisto- 

 gamic flowers of Ononis columnce the ten filaments are united 

 together into a tube. Again, in the extremely heterocleistogamic 

 flowers of Impatiens noli-me-tangere, I do not know on what 

 other theory to account for the fact, insignificant as it may seem 

 in itself, to which I have already alluded, that the brown cap-like 

 calyx has occasionally a protuberance on one side, which is evi- 

 dently a rudimentary nectary. If, however, this hypothesis is 

 correct, it is very remarkable how comparatively rarely we find 

 intermediate forms between the perfect and the closed flowers in 

 these species. 



The next point which attracts our attention is the large num- 

 ber of organs which have been correlatively modified in the ex- 

 treme heterocleistogamic flowers. Take, for instance, any of the 



