212 EEVIEWS. 



tinguishecl geuerically under the name o^ Krasclienihoivid), "floribus 

 superioribus sterilibvis, infimis (radicalibus) anantheris fertilibus 

 carnosulis." The flowers from the axils of the lower leaves become 

 buried in the soil and are described as " floribus * * clausis * * * 

 petalis staminibus stylisque nullis, capsulae rotiindatae parietibus 

 carnosis, seminibus fuscis * * * embryone peripherico arcuato 

 albumineque normali ! donatis." The normal flowers are petaloid 

 with the stamens nearly equalling the sepals. He says, " verosi- 

 millime capsulae intra imniculam steriles." M. Weddell* and Asa 

 Gray,t describe dimorphism in the genus Impatiens : M. Weddell 

 in the common I. Noli-me-tangere. In this plant some of the fruits 

 ripen Avithout the previous expansion of the flowers to which 

 they belong. All the whorls of the flower exist, but excepting the 

 ovary, they are extremely small and rudimentary, uniting into a 

 little hood, which the fruit, in elongating, bears up with it and wears 

 as a cap. These abnormal flowers arise near the normal ones, but 

 usually in lateral peduncles. Dr. Grray gives some interesting parti- 

 culars respecting the structure of the normal flowers of the American 

 species, in which certain membranaceous appendages of the filaments 

 are connivent and more or less coherent over the summit of the 

 pistil, entirely preventing the access of pollen in the greater propor- 

 tion of even fully developed flowers, which, consequently, fall away 

 unfertilized. In some, however, the growing ovary pushes the stigma 

 through the cap, thus securing its fertilization. 



M. Jussieu records dimorphous flowers in the section Meioste- 

 mones of the Natural order Malpighiaceae. In Acanthaceae {BueUid) 

 it was long ago observed by Dillenius. And we might adduce other 

 instances, but these must sufiice, for we possess no instance of this 

 kind of dimorphism, referred to our second category, which has been 

 fully and satisfactorily described, much less explained ; indeed the 

 examples which Ave have given are amongst the most marked and the 

 best observed. 



The main feature and that to which we would wish to direct 

 attention in, at least some, of these cases, is the occurrence of a 

 second kind of flower in which it would seem that nature has espe- 

 cially contrived to exclude the possibility of fertilization by other 

 thau own-flower stamens. It is true that the anthers in the closed 

 flowers of Viola and Oxalis are stated never to have been found open, 

 but in the Campanula observed by us the pollen evidently had access 

 to the stigma ; and indeed, M. Michalet points out, as we have said, 

 the existence of fijie threads coimecting the anthers with the 

 stigmas in the " hermetically closed " flowers of Oxalis. These fine 

 threads, there can be no doubt, are the pollen-tubes. It is 

 impossible that we should here enter upon the role of these remark- 

 able flowers in the economy of the species to which they belong. 

 We do not possess, as we have already said, a sufiicient basis of 



* Jussieu, ilalpighiaccesi, 85. f Gcu. United Statet-, ii. lol. 



