Geraniales.] 



BALSAMINACE^. 



491 



inuennost lobed pieces as two pairs of petals of a four-leaved corolla. A third view is 

 that of BeiTihai'di, who regards the exterior scales as bracts, the calyx as consisting of 

 five parts, of which three only, namely the spur and the back piece, which is double, 

 are present, and the others rudimentary or missmg ; while the corolla also consists of 

 five parts, of which the fovu' lower are united in pairs into the two innermost lobed 

 pieces, and the fifth is either separate, as in Hydrocera, or consolidated with the two 

 back mi.ited sepals into what he calls a petal-sepal. A fourth view is that of Kunth, 

 who considers the large back piece of the flower to be composed of two sepals, and 

 together with the spur and exterior scales to fomi a five-leaved calyx ; while he finds 

 in the two innermost parts a corolla of four petals united in pairs, and he assumes the 

 fifth petal to be abortive. This opinion has been adopted by Arnott in 1833, and by 

 Presl in 1836, the latter having discovered the fifth or missing petal to be present occa- 

 sionally in the garden Balsam, and always in Hydrocera triflora ; both these Botanists 

 finding in the genus Hydrocera the back piece, which is simple in Impatiens, composed 

 of two parts, and therefore confirming the accuracy of the theory of Kunth. Other 

 opinions, more or less resembling these, have been formed by others, but it is cleai' that 

 Kunth's theory is the only one that is correct. 



If we make a section horizontally through a young flower-bud of Impatiens macro- 

 chila, we flnd the following structm'e : — There is in the centre an ovary of five cells ; 

 with these alternate the five stamens, of which the fifth or anterior has a longer filament 

 than the others ; so far the structm-e is regvdar, and we have all the necessary evidence 

 of the flower, however irregular, being foimed upon a quinary type. Right and left of 

 the stamens stand the two innermost pieces ; these cannot be simple, because they are 

 opposite the intennediate stamens ; but their two-lobed figui'e, w'hen fuU grown, shows 

 that each is double, and then, their apparent centre being in fact their united margins, 

 they alternate with the anterior stamens, and so fall into the place usually destined for 

 petals. The last mentioned parts are half enveloped by the back piece, which might, 

 from its position, be the fifth petal ; but the case of Hydrocera showing it really to con- 

 sist of two united parts, they must be opposite the stamens, and consequently are sepals. 

 Next comes the spur, wliich overlaps the back piece, and stands opposite the antei'ior 

 stamen ; as no tendency to di\nde on the part of this piece is ever found, it must be a 

 sepal. Finally, the external scales, placed right and left of the whole flower, alternate 

 with those parts already sho^vn to be sepals, and consequently are recognised as the two 

 parts of the calyx required to complete the quinary plan of the whole flower. It will 

 be remarked, that a fifth petal has not been found ; if the eye is turned upon the back 

 piece, already foimd to be composed of two sepals, it mil be seen that a part is missing 

 between those two and the two coirresponding stamens ; and this is the place where the 

 abortion of a fifth of the corolla may, upon the evidence of this flower, be assumed to 

 occur; and where it is proved to take place by the evidence of Hydrocera, in wliich the 

 part missing in the Balsam makes its appearance. 



The annexed diagram ^^'ill serve to 

 illustrate the preceding observations; 

 the parts of the flower, as they really 

 exist in Impatiens being projected 

 upon a plane consisting of five circles, 

 of which the exterior (S) represents 

 the sepals or calyx, the next (P) the 

 petals or corolla, the third (s) the 

 stamens, the fourth (c) the carpels, 

 and the central (p) the placenta, or 

 axis. 



Connected with these plants is a 

 point of structure deserving of atten- 

 tion. In some species it will be 

 found that the style is surrounded 

 below its apex by five points, which 

 are evidently continuations of the 

 backs of the five carpellary leaves, 

 which certainly in these plants are 

 separate from the placenta, and are 

 merely pressed down upon it so as 

 to cover the OAoiles, thus confirming 

 the accuracy of the views concerning 

 placentation held by Schykofsky and 

 Schleiden. If so, what else can the 

 upper part of the style and the stigmas be, except the naked apex of the placenta, 



