44 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN 



petaloid. They may be but slightly enlarged and spreading 

 under the capsule or appressed to it, or markedly accrescent and 

 completely enclosing it, or even extending far beyond it like a 

 pair of valves. 



The corolla is always strongly two-lipped, and spurred or rarely 

 only saccate at the base of the lower lip, but otherwise shows great 

 diversity. Each genus possesses its own characteristic type of 

 corolla, easily recognized at a glance when one has become familiar 

 with it, but sometimes difTficult to define in words. The palate is 

 very different in different genera; two-lobed in some, nearly 

 hemispheric in others, helmet-shaped in Stomoisia; in Calpidisca 

 the limb is strongly reflexed from the aperture of the spur, the 

 palate consisting of a more or less thickened and sometimes 

 toothed ring along the line of reflexion; in Vesiculina the palate 

 is a prominent and divergently two-lobed sac involving essentially 

 the whole of the two lateral lobes of the lower lip; in Lecticula the 

 palate is nearly obsolete and the corolla is almost that of Pingui- 

 cula. 



One of the noteworthy family characters commonly assigned to 

 the Lentibulariaceae is the possession of one-celled anthers. It is 

 probably true that the anthers are never two-celled with the cells 

 collateral, as is commonly the case in related plants; but through- 

 out the group with bractlets, the anthers are always more or less 

 strongly transversely constricted and consequently partially if 

 not completely vertically two-celled. This anther-structure de- 

 serves special study, particularly from the standpoint of its 

 developmental history. So, too, does the very peculiar anther- 

 structure of a small group of South American species, here referred 

 provisionally to Calpidisca, but probably worthy of generic rank 

 when better understood. In these, there is developed at the 

 summit of the filament a broad horseshoe-shaped membrane, 

 called for want of a better term a "connective," bearing on its 

 inner edge two anther-cells, or perhaps a single strongly con- 

 stricted one. 



The ovary, placenta, and stigma are very similar throughout 

 the family. So also is the capsule, although the method of 

 dehiscence may prove of some value in classification when known 

 for all species. The seeds present great variations, but these are 

 chiefly in the form and markings of the outermost seed-coat. 



