THE PER TILISA TION OF FL WERS. 335 



obtained from a previously visited dead-nettle, shall be deposited 

 on the stigma. The stigma, as we have seen, depends below the 

 anthers. Hence it must be the first object with which the bee 

 comes in contact. Fertilisation by the foreign pollen is thus secured 

 before the stamens have dusted the insect with the flower's own 

 pollen. As Dr. Ogle has pointed out, the position of the stamens 

 doubtless facilitates in a marked degree the proper placing of pollen 

 on the insect's body. If the anthers had lain side by side, the bee's 

 head might have been dusted on parts which do not touch the 

 stigma as the insect enters the flower ; whilst even the eyes of the 

 bees might have become disadvantageously covered with pollen. 

 There is, in short, the closest possible correlation between the structure 

 of the flower and the form and size of the insect which fertilises it. 



Such correlation is exhibited in, if anything, an increased degree in 

 the genus Salvia, belonging to the dead-nettle order (Labiatce), also 

 including the sage (Fig. 235). Salvia (Figs. 231, 232) attracted the 

 notice of Sprengel Rector at Spandau who, in his *' Das entdeckte 

 Geheimniss der Natur im Bau und in der Befruchtung der Blumen" 

 ("The Secrets of Nature in the Structure and Fertilisation of Flowers"), 

 published in 1793, was one of the first to direct attention to the fact 

 that nature's law was " cross " and not " self-fertilisation " or, as he 

 himself expressed it, " nature does not desire that any complete flower 

 should be fertilised by its own pollen." It is interesting to note that 

 a species of Salvia (S. splendens) occurring in the New World 

 appears to be cross-fertilised through the agency of humming-birds ; 

 these fairy-like birds thus discharging in this case the functions of 

 the insects to which some species approach so nearly in size. The 

 trumpet creepers (Tecoma radicans) and trumpet honeysuckle 

 (Lonicera sempervivens) are probably fertilised by moths and by 

 humming-birds as well. 



In Salvia officinalis (Fig. 231, A), the general form of which 

 closely resembles that of the dead-nettle, the stamens ripen before 

 the pistil ; and as, moreover, the stigma (/) is placed above the 

 anthers (a), self-fertilisation is an impossibility. When, however, the 

 stamens have shed their pollen, they shrivel up, and the pistil as 

 it ripens develops the stigma, so that it elongates, curves down- 

 wards, and thus assumes a position (Fig. 232, B, st) in which it 

 cannot escape contact with the back of the bee entering the flower 

 (Fig. 232, A). The insect's back, it may be noted, is exactly 

 that region which the ripe stamens in a younger and necessarily 

 different flower will have dusted with foreign pollen. But the 

 economy of Salvia includes yet other appliances for more effectually 

 securing fertilisation by the insect. There are but two well-deve- 

 loped stamens (Fig. 231, B) in the flower. These organs have 

 widely separated anther-cells (a, a') ; and when in an undisturbed 



