52 PITTONIA. 



with segments distinct, others with more or less conjoined 

 segments ; besides some with petals like the sepals, and some 

 with petals and sepals very dissimilar in size, form or atti- 

 tude. In connection with the androeciura one may note 

 filaments wholly free from the perianth in most species, 

 while in not a few they are more or less ad n ate ; and there 

 are filaments filiform throughout ; filaments toothed below ; 

 filaments subulate- or even triangular-dilated ; filaments 

 widening upwards ; filaments petaloid-dilated, tricuspidate at 

 the apex, the central cusp antheriferous ; and there are fila- 

 ments cohering below, thus forming a second crown or tube 

 within the perianth. And such differences as these are, else- 

 where in the order, regarded as amply sufficient for the 

 establishment of separate gencTa. In view of differences^ 

 such as some of these, all botanists anterior to Haller, and 

 many in more recent times, considered the Leeks, Onions and 

 Garlics as of distinct genera ; recognizing species of Forrnm, 

 Cepa and Allium. Even Linmeus, not given to the accepting 

 of genera without good characters, admitted these three with- 

 out any hesitancy, until Hnller taught that they could consti- 

 tute but one natural genus." But to Haller's view, even after 

 it had been adopted by Linnseus, there Avere eminent authori- 

 ties who could not subscribe ; and a long succession of men 

 thus minded, was continued almost down to our own time. 



At least three hundred species of these plants are now 

 known ; and out of them have been selected the types of 

 more than twenty proposed genera. The authorship of such 

 proposed genera rests with botanists of such high standing 

 as Tournefort, Adanson, Medikus, Moeuch, Salisbury, Lindley, 

 Kunth, Eeichenbach, Parlatore and several others of equally 

 recognized ability. All these have seen, in the fusion of so 

 many and such diverse types into one genus, a prononnced 

 contradiction of those morphological principles on which all 

 classification rests, or is presumed to rest. 



* See /I. dkhhunydeum and crisfuiu. Pitt. i. 166. 



'■' De Allii Genere Naturali Libellus. GcEttingfe, 1745. 



