262 On the Fertilization of the Blue Lobelia. 



with the pollen. When an insect leaves one flower and goes 

 to another, possibly a flower lower in position and in a later 

 stage of development, on another stem or plant, he will very 

 likely find the stigma of that flower expanded and protruded. 

 If so, it will just sweep his back dusted with the pollen of the 

 previous flower ; and if he then mounts to a flower higher ir^^ 

 position and in an earlier stage of development on the second 

 stem or plant, he will probably find the anthers in a state to 

 give up their pollen to him, and so on. 



The flowers are very commonly out two at once on one 

 stem, the lower one with the stigma protruded and unfolded, 

 and the upper one with the stigma still within the anther- 

 tube, and the anther-tube ready to discharge its pollen at a 

 touch. The number of flowers visited by a humble-bee in a 

 few minutes is very remarkable. 



It is interesting to watch the gradations of this curious 

 structure in GaTnjpanula^ Jasione^ and Lobelia. All have the 

 stamens set upon the calyx ; all have the brush on the style 

 for sweeping out the anthers ; in all the stigmatic surface 

 remains closed until the pollen has been swept out ; and in all, 

 when the stigma opens it turns its back on its own pollen, if, 

 indeed, any is then left on the style. But there are great 

 differences. 



In Camjpanula the brush is long and the three-lobed stigma 

 large ; the pressure of the growing style with its brush against 

 the anthers is effected by means of the edges of the lobes of 

 the valvate corolla, which are folded deeply inwards, and, 

 being opposite to the stamens, press against their backs as the 

 flower opens ; and the transportation of pollen must be effected 

 by the moving about of the insect within the bell and against 

 the pollen-covered style. There must be a profuse expendi- 

 ture of pollen ; but even here there seems to be a wonderful 

 economy in the bristles, which are not scattered on the style 

 promiscuously, but are set on in ten rows apparently opposite 

 the ten anther-cells. 



In Jasione the brush is shorter, and the two-lobed stigmatic 

 surface quite small ; the long thin lobes of the corolla do not 

 press the anthers against the brush, but, instead, the anthers in 

 the opened flower are syngenesious. Their attachment to one 

 another, however, is slight, is at the base only, and does not exist 

 in the early bud. The transport of pollen must be effected by 

 the insect walking over the numerous flower-heads, and 

 amongst and against the long protruding styles, of which 

 some in each head are generally pollen-covered, with closed 

 stigmas, and others pollen-stripped, with opened stigmas. 



In Lobelia the anthers are sliort, the brush on tlie style is 



