Constricted Entrance to Nectar-cavity. 113 



plan, by which the perianth is made to serve the same 

 purpose as when provided with these straight and 

 twisted passages and spurs, is its constriction at some 

 or other spot. This may be seen in many Asperifoliae 

 and Primulacese, for instance in the several species of 

 Androsaceae, where it is the corolla that is thus con- 

 stricted. (Plate II. fig. 46, longitudinal section of 

 flower of Androsace glacialis.) 



In the same category must be classed the swellings 

 and tubercles that occur in such variety on the perianth, 

 round about the entrance to the nectary. These pre- 

 sent such an endless multiformity that it is impossible 

 here to give an exhaustive account of them. And this 

 so much the more, as the attempt to describe them 

 would in each case involve a detailed consideration of 

 other correlated structures in the flower. I shall there- 

 fore merely select a couple of typical cases out of the 

 vast mass of material, cases which I have attempted to 

 illustrate by drawings ; and even in dealing with these 

 I shall limit my remarks to the form and significance 

 of the tubercles and swellings, without discussing any 

 other structural arrangements which may be correlated 

 with them. 



One of the simplest arrangements is that of which 

 Nigritella angustifoUa may serve as an example (Plate II. 

 fig. 48, front view of flower ; fig. 47, longitudinal section 

 of fiower,in which the tips of the perianth were deficient). 

 The nectar is here secreted by a saccular pouch of the 



