﻿factorily, the bilabiate corolla, the palate, the lower lip's midlobe 

 provided with a number of great reflexed setae, and the four 

 stamens — their filaments united in pairs below. In only one point, 

 and that of secondary importance, is there any difference, viz., in 

 the presence of six ovules in each ovarian cell, instead of four. But 

 when one remembers the affinity of Mel leva, this is only what would 

 be expected ; indeed, it seems to emphasize the difference between 

 Mellera and Galophanes with its couple of ovules in each cell. 

 Exception might perhaps be taken to this statement on the score of 

 the inflorescence — which at first sight seems very different in the 

 two species. But after all, it is only the difference between a spike 

 and a narrow panicle, dimorphic flowering branches being known in 

 other genera, comprising also species not provided with such — of 

 which a good example is the genus Petalidium. 



Besides the inflorescence and ovules, the following are the chief 

 1 M. lobulata and M. nyixx.nut ; the leaves of the 

 ntly shaped and not lobulate, the upper lip of its 

 still more deeply 2-lobed and the lobes are wider, 

 nished with more numerous setae, and, judging 

 is more prominent. The anthers also 

 is different. 



flower are to be seen beautiful adaptations 

 About the middle of the abaxial side of 

 uple of small pouches, into which probably 

 3 applied, the two dilated bases 

 of the conjoined filaments. This arrangement, together with the 

 setae on the lower lip and the spurs to the anthers, seems well 

 calculated to direct an insect's thrust, and at the same time cause 

 much movement on its part while visiting the flower, thus raising 

 the chances of pollination. 



So much attention has been paid of recent years to the pollen of 

 Kadlkofer,* and lately by Dr. Lindau,t 



Acanthacea, first bv Professor Kadlkofer,* and lately by 

 that a few words on the pollen of M. nyassana may n< 

 place. Like that of M. U>bulata,\ this pollen is of the kind called 

 by our German friends "Bippenpollen," which in its typical form 

 Lindau finds to be characteristic of the tribes Hygrophilea and 



ellipsoidal than those of M. lobulata, and they have fifteen closely- 

 placed ribs, there being four ribs between each pair of pores, 

 while each pore is situated on a rib ; the lips of the pores are very 



