130 THE VICTORIAN NATURALIST. 



found in a flower. They are imprisoned, and cannot escape 

 because, as mentioned, the throat of the flower is beset like a 

 trap with long motile hairs, which bar the passage out as in a 

 weir basket. While the insect is thus wandering around the 

 cavity it brings its pollen-laden back in contact with the sti^matic 

 surface and pollinates it. As soon as this has taken place, the 

 anthers, which have been closed hitherto, dehisce and become 

 freely accessible at the same time by the change in the stigma. 

 About this time the hairs in the narrow passage wither and 

 collapse, leaving a free exit, and the midges, all be-powdered 

 with pollen, hasten to leave the flowers. That they retain no 

 unpleasant recollection of their temporary confinement may be 

 inferred from the fact that they have no sooner escaped from one 

 flower than they creep into another. 



It may be mentioned that so long as the stigma is still re- 

 ceptive, the perianth opens outwards, presenting to the flies a 

 hospitably open door ; but as soon as they have accomplished 

 the pollination of the stigma, and when the flies, again laden with 

 pollen, have flown away from the flower, the banner-like lobe of 

 the corolla closes over the mouth of the throat, stopping the 

 entrance to the flies, which have now nothing more to do here. 

 In fact, it were a contradiction for the invited guests on their 

 arrival that a flower should remain widely open when no more 

 nourishment was to be obtained — when the meal, so to speak, 

 was finished. VVhen this is the case it is undesirable that they 

 should interfere with the other younger flowers by competing with 

 them for visitors. 



In very many of the Papilionaceae the two lateral petals, called 

 wings, converge towards their upper margins, along which they 

 are in contact, so that they form a convex saddle, arching over 

 the keel. The wings and keel are locked together, and every 

 pressure upon the pair of wings is transmitted to the keel, in 

 which are the anthers. Consequently when an insect sets itself 

 astride on the saddle-ridge formed by the wings, not only is the 

 latter pressed down but also the keel, and this movement is 

 accompanied by the extrusion of a pasty vermicular mass of 

 pollen, and by the simultaneous adhesion of the pollen to the 

 insect's belly, or sometimes to its legs. In this case the action 

 of the stamens within the cavity is just the same as that of the 

 piston inside a pump, and has therefore been called "pump 

 apparatus," which appears to be confined to papilionaceous 

 flowers only. 



The mechanism to be described next does its work by means 

 of impact, and performs the movement of the filaments resembling 

 the striking of a hammer. The best known examples of the 

 hammer form of mechanism occur in the genus Salvia. The 

 under lip of this labiate flower serves as a landing stage for the 



