294 THE CROSSING OF FLOWERS. 



stamens and pistil, and so might also be mistaken for hermaphrodite flowers, but 

 closer examination shows that their ovaries do not develop sufficiently to produce 

 fertile seeds. The ovules, and usually the stigmas as well, do not mature, but the 

 pollen in the anthers attains its full power. (3) True pistillate flowers. In these 

 only fertile ovaries are developed, and there is no trace of stamens. (4) True 

 staminate flowers, the counterpart of the third group. They contain stamens whose 

 anthers hold ripe pollen, but their ovaries are quite suppressed. 



To the four types of unisexual flowers we shall apply short descriptive names, 

 at once rendering the character of the flower apparent, and saving much needless 

 iteration and confusion. (1) Pseudo-hermaphrodite feTriale floivers are the pistillate 

 flowers which appear to be hermaphrodite, similarly (2) Pseudo- hermaphrodite 

 "male flowers are the staminal flowers which, from the presence of reduced pistils, 

 appear to be hermaphrodite. (3) True pistillate flowers, and (4) True staminate 

 flowers. 



Next to the unisexual come the neuter (sterile) flowers, in which the ovaries and 

 stamens are either altogether absent, the flower consisting merely of petals and 

 sepals, or the sexual organs if present are quite rudimentary and hidden away in 

 the centre of the flower. 



The kinds of flower enumerated here are connected together by numerous 

 transitional forms. In the hermaphrodite flowers of the Knawel (Scleranthus) two 

 or three of the four stamens are often sterile; they occupy their right position, but 

 their anthers are shrivelled and contain no ripe pollen, only one or two of the 

 stamens being properly developed. Of the eight stamens of the well-known garden- 

 plant Clarkea p)idchella only the four which alternate with the petals form fertile 

 pollen, while the anthers of the other four are abortive. Sometimes Ave, six, 

 seven, or even all the anthers are sterile. The Chickweed {Stellaria media) has 

 ten stamens arranged in two whorls of five, but it rarely happens that all the 

 anthers produce fertile pollen. Usually those of the five inner and often of a pair 

 of the outer whorl are shrivelled and have no pollen. These instances evidently 

 form good links between the true hermaphrodite and the pseudo-hermaphrodite 

 flowers. The flower-heads of the Burnet {Poteriwni polygamum) consist of pistil- 

 late, staminate, and truly hermaphrodite floM^ers. In the staminate flowers sixteen 

 stamens are formed; the hermaphrodite flowers may contain eight, seven, six, or 

 gradually decreasing numbers down to only one. The other stamens are not even 

 formed, not the slightest trace of them can be found. These flowers may be 

 regarded as connecting the truly hermaphrodite with the pistillate flowers, for if 

 the suppression of stamens be supposed to go still further, so that the last stamen 

 has disappeared, then the flower is no longer hermaphrodite, but has become a true 

 pistillate flower. 



The gradations in the class of pseudo-hermaphrodite, pistillate, and staminate 

 flowers are also very varied. The Fuller's Thistle (Cirsium), the Flowering Ash 

 (Fraxinus Ornus), the Asparagus (Asparagus offixinalis), the Date-plum {Diospyros 

 Lotus), the Vine {Vitis vinifera), many Scabiouses, Saxifrages, Valerians, &c., all 



