2X0 KEVIEWS. 



hypogean flowers of Linaria spuria which ripen seed, but these 

 flowers offer no notable difference from the rest. 



We have ourselves examined the dimorphic flowers both of Viola 

 and Oxalis, but have nothing to add to M. Michalet's detailed 

 account. Another case of similar character, however, in a far 

 removed natiu'al order, has recently been closely under our notice ; 

 and as it is one which we believe is not widely knowTi, although it 

 has been carefully investigated by M. Brongniart, and is described 

 by M. Ad. de Jussieu in his " Monographie des Malpighiacees," 

 and by Torrey in his 'Flora of New York' (i. 428), vce may be 

 permitted to give some account of it here. Our observations 

 have been made solely upon dried specimens, so tliat we can only 

 speak to structural facts. The most important problems which 

 these phenomena suggest, it is needless to say, can only be solved 

 by study and carefully watching of the living plant, and this we have 

 not yet had the opportunity of doing. Linnaeus long ago* wrote of 

 Campanula perfoliata— " flores laterales raro corollo instructos pro- 

 ducit, sed tantum calycem semen continentem ; flores vero terminales 

 perfecti sunt." This species we have not examined, though, from 

 M. Jussieu's accountjt it would appear to have been the same that 

 M. Brongniart observed. This excellent botanist found, in the 

 smaller flowers, a 'tympanimi' covering the base of the calyx-tube. 

 This is the rudiment of the corolla. On removing it he found the 

 stigma and stamens with well-formed pollen. Our attention has 

 been devoted to two Indian species (C. canescens) alluded to by 

 Messrs. Hooker and Thomson in their Prsecursores ad Floram 

 Indicam,! and G. coforata. We find the smaller and apparently 

 apetalous flowers to be usually lateral, often borne on short racemes 

 springing from the lower leaves of the stem, though sometimes they 

 abound along the branches of the principal inflorescence. They are 

 various in size, from that of a coriander seed to a pea. The ovary is 

 inferior, as in the normal flowers, though occasionally only two-celled. 

 The limb of the calyx is not always regularly five-lobed, but often, 

 and especially in the smallest flowers, three-lobed or irregularly 

 divided. The disc of the flower is covered by a completely closed 

 hairy membrane, with a slight mammilliform elevation in the centre. 

 This membrane is the rudimentary corolla, and the number of petals 

 composing it is indicated by converging lines. On dissection, this 

 elevation is found to cover a pentagonal or five-lobed body which at 

 first sight resembles a stigma, but examination shows that each lobe 

 is opposite to a corresponding lobe of the calyx-limb, and that each 

 lobe of the pentagonal process is united to the base of the opposed 

 calyx-lobe by a delicate cord, the filament in fact of the stamen, of 

 which the lobe of the central body to which it is imited is tlie anther. 

 The stamens are extended horizontally between the upper mem- 



* rrtclcctioiics Botauicw, 'J99. f Op. ell. 84. 



:j: .Jouni. Liuu. Soc. ii. 7. 



