258 Dr Dickson on Diplostemonous Flowers^ (&c. 



petals is only an apparent exception to the rule I have 

 stated above, as to the position of the carpels in 5-carpellary 

 MalvacesD ; for the researches of Payer on the organogeny 

 of Pavo7iia leave no doubt that in this tribe the five loculi 

 merely represent the fertile members of a circle of ten 

 carpels, to which the ten styles correspond. In Favo7im, 

 the gynoecium, in its origin, consists of ten carpellary mam- 

 millae. Of these, however, only five have loculi developed 

 in connection with them — every second carpel being, so to 

 speak, barren. The ten carpels all equally develope styles ; 

 so that in the advanced condition there are five styles pro- 

 longed upwards from the loculi, and five continuing the 

 lines of the dissepiments.* In Malvaviscus, Payer describes 

 the loculi (corresponding to the fertile carpels) as superposed 

 to the petals. t In Urena {U. americana, U. lobata, U. 

 scahriuscula, U. sinuata), I find the same arrangement. In 

 Pavonia, 1 have as(^ertained the remarkable fact that the 

 loculi are sometimes superposed to the sepals, and sometimes 

 alternate with them. Thus, in P. typholea, P. hegonicefolia 

 (Gardner), P. odorata, P. iimbellata, and P. zeylanica, the 

 loculi are superposed to the sepals ; while, in at least one 

 species, named in the Edinburgh University Herbarium 

 P. hastata,X the loculi are certainly alternate with them, as 

 in 3Ialvaviscus and Urena.% 



^ Organog6nie, p. 35, pi. 7. 



t Le9ons sur les fam. nat. des plantes, p. 281. 



X I have expressed myself thus guardedly as to the specific name of this 

 plant, because, by its indefinite stamens, it differs from that to which Baillon 

 refers as P. hastata (Adansonia, II. p. 176), which is described by him as 

 having only five stamens in the adult state. The Edinburgh plant agrees 

 with the description of P. hastata in Decandolle's " Prodromus" (vol. i. p. 443), 

 in its lanceolate hastate dentate leaves, axillary unifloral pedicels, and five- 

 leaved involucre. I cannot say much as to the colour of the petals, except 

 that a deep red or purplish blotch remains at the base of each. The whole 

 plant (especially the stem, the under side of the leaves, the involucre, and the 

 sepals) is downy, being covered with a short stellate pubescence. The plant 

 which Payer has examined as P. hastata appears to perfect a considerable 

 number of stamens, as is seen in his representation of the androecium " shortly 

 before blossoming," where there would seem to be 25 stamens, or there- 

 abouts (Organogenic, pi. 7, fig. 9 ; with description, p. 38). 



g The position of the fertile carpels seems to offer a much more important 

 character by which the genus Pavonia may possibly be disintegrated, than any 

 derived from the awned or awnless condition of the fruit, the relative length 

 of the involucre to the calyx, &c. 



