no Flowers and their Unbidden Guests. 



expenditure of the nectar; or they compel those 

 visitors whose size is suited to the mechanism of the 

 flower, and who would therefore conduce to allogamy, 

 to make their entrance in the exact way by which this 

 advantageous result can be insured. The formations 

 in q^uestion may therefore in this latter case be classed 

 with those thorns and hair-Hke trichomes, which we 

 designated as " path-pointers " in the two last chapters. 



In this constriction of the passage which leads to 

 the nectar we find again the same variety as to the 

 methods employed as when the nectar-cavity is closed 

 completely. Sometimes it is effected by tubercles, 

 gibbosities, and various deformations of the perianth, 

 or of excrescences of the perianth ; sometimes by the 

 crowding together of some or other of the floral organs ; 

 sometimes by the interposition of stout ovaries ; some- 

 times by the thickening and dilatation of the filaments. 



One method which repeatedly recurs, and of which 

 Narcissus juncifoUus (Plate II. fig. 57, longitudinal 

 section of flower) and (Enothera grandifiora (Plate II. 

 fig. 51, longitudinal section of flower) may be selected 

 as examples, is the contraction of the lower part 

 of the perianth to one or more narrow channels. 

 These however, though narrow, are yet wide enough 

 to allow large insects to pass their long thin trunks 

 down to the nectary, while smaller insects with 

 shorter proboscis are completely excluded. In one 

 of the two flowers selected as examples {Narcissus 



