Exclusion of Ants. 23 



this it is easy to convince oneself. Pluck, for instance, 

 some of the flowers of Melianthus, or of any other plant 

 in which, as in this, the nectariferous flowers, whUe 

 growing in their natural position in the inflorescence, 

 are admirably protected against the visits of ants, and 

 having plucked them lay them on the ground. They 

 are now unprotected, and in the shortest possible time 

 they wiU be found swarming with ants. Another 

 example is furnished by Phygelius capensis. Here the 

 flowers are rich in nectar, all access to which is rendered 

 impossible to ants during the period of flowering, by 

 a method which I shall have to describe later on in 

 detail. So soon, however, as the corolla detaches itself 

 from the torus, the nectar, of which there is still an 

 abundant store, becomes readily accessible, and ants 

 (in the Botanical Gardens at Innsbruck abundance of 

 Lasius niger) immediately crowd in and greedily lick 

 it up. They can be allowed to do so at this period 

 without iU result. Tor, now that the flowers are falling 

 ofi', their nectar is useless, and no longer wanted to 

 attract such flying insects as cause intercrossing. 



In this respect the case of Antirrhinum maj'us 

 L. is also very instructive. Here the corolla remains 

 closed so long as the stigma is unfecundated. Strong 

 humble-bees can indeed force an opening and so bring 

 about intercrossing; but ants are quite unable to 

 squeeze in between the two closely compressed lips. 

 But, when once the stigma has been covered with 



