252 TAKING UP OF POLLEN BY INSECTS. 



is not accessible direct, but is concealed in tubes or recesses, and the covering must 

 be removed before an insect can be besmeared. In the composite flowers of the 

 genera Onopordon and Gentaurea, to which belongs, amongst others, the well-known 

 Corn-flower (Gentaurea Gyanus), the anthers are borne on slender filaments, and, 

 as in all Compositse, are connate into a tube, in which is concealed the upper portion 

 of the style. The dehiscence being introrse, the pollen is deposited on the style. 

 In the majority of Compositse, the style then grows in length and pushes the pollen 

 up beyond the top of the tube. But this is not the case in Onopordon or Centav/rea. 

 No elongation of the style occurs, and the pollen remains concealed in the tube. 

 If, however, an insect sets foot on the ceritral part of the capitulum and comes into 

 contact with the stamens as it clambers over the florets of the disc, the filaments 

 immediately contract, drawing back the sheathing anther-tube and leaving the 

 pollen exposed on the top of the style, which is then brushed against by the under 

 surface of the insect. The same result is achieved by diflerent means in certain 

 Papilionacese. The well-known Gytisus, Melilotus, TrifoUum, Onobrychis are 

 instances of one group. The front pair of petals, which is known as the keel, and 

 serves as an alighting-place for insects, forms a receptacle with a very narrow 

 opening at the top. In this are concealed the ten stout, partially connate filaments, 

 and the anthers borne by them. When a humble-bee settles on the keel and inserts 

 its proboscis into that part of the flower where honey is to be found, the keel is 

 pressed down by the insect's weight, and the anthers are in consequence exposed, 

 whilst the pollen resting upon them is rubbed ofl" on to the under surface of the 

 insect. The moment the insect leaves the flower the keel springs back to its former 

 position, and once more conceals the anthers which, as a rule, have only parted with 

 a portion of their pollen. The same process is repeated on occasion of each fresh 

 visit, and as many as four difi^erent insects may thus be dusted with pollen from 

 the same flower. In Lathyrus, Orobus, Pisum, Vicia, the phenomenon is in the 

 main the same, but these plants have a special brush developed in connection with 

 the style which sweeps the pollen out of the keel, where it has been deposited by 

 tlie anthers, at the same moment as the insect alights on the flower. The insect is 

 thus certain to carry away pollen on the under parts of its body. 



The transference of the pollen in the Hemp Nettle (Galeopsis) and Monkey- 

 flower (Mimulus) to the bodies of insects is also attended by a curious phenomenon. 

 A stamen of Galeopsis is shown in fig. 216 ^^, p. 91. The corolla is bilabiate, and 

 beneath the arch of the upper lip are two pairs of stamens, one pair behind and 

 the other in front of the stigma, which is bilobed, and has the property of shutting 

 together its two component flaps in response to contact. Each of the anthers of the 

 pair behind the stigma is box-like, and divided by an internal septum into two 

 compartments capped by lids. If a needle is inserted into the flower so as to touch 

 the anthers, the lids spring open, disclosing the pollen, which sticks to the needle; 

 a similar action occurs when an insect visits the flower. 



These cases, where the pollen has first to be uncovered by the insect before it 

 can be carried away, are not more remarkable than those wherein pollen-masses 



