AND THEIR SYSTEMATICAL ARRANGEMENT. II9 



out a pedicle, and of such fine grains, that they 

 resemble a farinaceous substance. In those of the 

 last division, with deciduous anthers, the pollen- 

 masses consift of round, and often divisible globules, 

 but never acquire the mealy appearance of the 

 former. The anthers of the second division, when 

 unripe, resemble indeed thofe of the third, but 

 never drop off. These circumstances seem to give 

 additional weight to the proposed arrangement. 



As all the genera I have given are founded 

 upon the results of my own observations, I did 

 not choose to take up the several new ones of the 

 Prodromus flora? Peruviance & Chilensis ; most of 

 them being known to me by figures only. Gongora, 

 Masdevallia, and perhaps Anguloa appear to be 

 distinct from the genera established by me, though 

 it is certain, that they all belong to the third di- 

 vision. 



