1 8 Mr. T, Smitli on certairi Species of 



I am aware, neverthelesss, that this is a very doubtful test 

 in a cultivated plant, the flowers of which are frequently bar- 

 ren from causes that are not obvious. 



By the kindness of Mr. Anderson I was enabled several 

 times to examine the numerous species of the genera Sen-atnla^ 

 Carduus, and Cfiicus, which are cultivated in the Botanic Gar- 

 den at Chelsea ; and about half the plants to which, fi-om the 

 state of their flowering, I could apply the tests above men- 

 tioned, proved either male or female. 



In Serrafida, the only species not hermaphrodite was the 

 tinctoria. 



In the genera Carduus and Cnicus I ascertained the follow- 

 ing, as named in Mr. Anderson's manuscript catalogue, to be 

 female plants. 



Cnicus tuberosus, ochroleucus, semipectinatus, Sc Salisburgensis. 

 Three or four others I suspect to be male plants ; for, upon 

 examining many capitula that had flowered, I could not find 

 any perfect seeds. 



I have looked over the specimens of Carduus and Cnicus in 

 the Banksian Herbarium, and the following appear to be fe- 

 male plants : 



Carduus rivularis, Chius, rigcns, serratuloides, paniculatus. 

 Cnicus leucocephalns, rigens^ Erisit/iales, tuberosus, acaulis, 

 oleraceus. 



There are specimens of both sexes o? Erisithales and acaulis; 

 the specimen of the female plant of acaidis is remarkably di- 

 stinct from the male. 



Since I first turned my attention to this subject, a doubt has 

 arisen whether in many, perhaps in most of the cases in which 

 female plants occur, the antheriferous plant may not be an her- 

 maphrodite rather than a male. 



The plant which I first ascertained to be dioecious was Cniais 

 arvensis .- in this the separation of the sexes is undoubted and 

 unequivocal ; for though I have examined a very great num- 

 ber of male plants, the ovaria have always proved abortive, 

 except in one instance, in which two of the ovaria in one capi- 

 tulum were most decidedly impregnated, tlie embryo being so 

 far advanced that no doubts could be entertained about it : the 

 stigmata of these flowers did not, however, appear to differ 

 fi'om those of the numerous unimpregnated ovaria which sur- 

 rounded them : this case must therefore be considered as 

 merely accidental. 



Having ascertained that this species was dioecious, I con- 

 cluded that all the others were so in which female plants were 

 to be met with ; but, in some, hermaphrodite plants certainly 



occur, 



