SPRINKLING APPARATUS. 271 



is followed by a transference of the pollen to the bodies of the insects. Unfor- 

 tunately we do not know of any observations having been made of the visits of 

 insects to plants of this species in the wild state. Such observations might enable 

 us to come to a sure conclusion on the subject, but so far all our results have been 

 derived from flowers reared in hot-houses. 



Next to the expulsive variety of mechanism comes the sprinJding variety. 

 Pollen transferred by this kind of apparatus is always of mealy or powdery 

 consistency, and is shaken out of the loculi where it is produced. Three 

 modifications of this apparatus may be distinguished, viz.: — the sugar-tongs 

 modification; that in which the anthers dehisce by terminal pores; and that in 

 which the anthers are united into cones. In contrivances of the first kind, 

 the filaments are like the arms of a pair of sugar-tongs, and the anthers borne 

 by them are, when in the dehiscent condition, in the form of spoon -shaped 

 receptacles or recesses, with the concave sides facing one another. Pollen of 

 mealy consistency would not stay in open, upright spoons of the kind were it 

 not for a special contrivance. In order to picture the state of things it is best 

 to think of the action of a pair of sugar-tongs in which the end of each arm is 

 fashioned into a spoon. When the tongs are closed the concave surfaces of the 

 spoons are brought together, and form a receptacle in which sugar can be 

 retained in the form of little solid bits, and even in the form of fine powder 

 if the parts of the tongs fit well. The moment the two arms of the tongs are 

 separated the contents held by the spoons drop down, and if in the condition 

 of powder they must inevitably bepowder any object that may happen to be 

 underneath. Now, this is just what occurs in connection with the sprinkling 

 apparatus in the flowers of a large number of Acanthacese, Ehinanthaceae, and 

 Orobanchacese. Beneath the protective covering of the floral envelopes — most 

 commonly under the upper lip of a bilabiate corolla — are found the stamens 

 arranged two and two, with the anthers of each pair closed together like the two 

 valves of a leguminous pod. They are kept in this position by the stiflT staminal 

 filaments, and the margins of the valves fit one another so exactly that not a single 

 pollen-cell from the mass of floury pollen contained between them can fall out 

 unless the structure is subjected to some particular shock. In some cases each pair 

 of opposite valves is furnished with matted hairs which join the upper edges 

 together. The form is then like that of a mussel-shell, and only the free margins 

 can open. The moment the valves separate in the least, whether they are joined 

 together at one part or not, the pollen spills out in accordance with the law of 

 gravity. The direction of its fall is often determined by delicate hairs fringing 

 the margin of the anther, the object being to prevent waste. The separation of the 

 valves is caused by insects — and in tropical regions probably by humming-birds as 

 well — when they press into the jaws of the flowers in search of the honey stored 

 in the back-ground. In doing so the proboscis (or beak) is either pushed right 

 between the valves, or it strikes against certain special processes with which the 

 valves are furnished, or else the tense filaments bearing the valves are forced 



