JOHNSON: A REVISION OF THE SECTION BORAPHILA 3 



respectively (but are distinctly characters of minor importance) the classi- 

 fication automatically tended towards a natural arrangement, the significance 

 of which Don seems not to have fully appreciated. 



Tausch (30) grouped the species in three sections, Micranthes, 

 Hydatica, and Arahisa. He, too, based his distinctions on the characters of 

 the calyx. Thus the calyx of Micranthes was described as adhering to the 

 ovary and reflexed ("calyx adhaerens, ore reflexo"), that of Hydaiica and 

 of Arahisa as free and reflexed {"calyx liber, reflexus"). With the excep- 

 tion of the section Arabisa, where the species are well defined, Tausch's 

 reliance on minor characters resulted in the grouping together of species 

 of widely different affinities. The characters of the carpels were entirely 

 overlooked. 



Seringe (26) followed the classification of Tausch, with its attendant 

 confusion, in the relationships of the species. Neither Seringe nor Tausch 

 appreciated the importance of epigyny and hypogyny, and Haworth's 

 accurate characterizations of the genera Robertsonia and Aulaxis seem to 

 have been completely ignored by them. Torrey and Gray (31) adhered 

 to the system of Seringe, although they somewhat stressed the epigeous 

 and hypogeous character of the caudex, a character which is subject to 

 variation under different ecological conditions and unreliable as a character 

 of major taxonomic consideration. Hooker ( 16) followed the same system 

 in Flora Boreali Americana. 



The system of classification proposed by Tausch and followed more or 

 less closely by the subsequent writers above mentioned was considerably 

 modified by Ledebour (18). He combined the sections Arahisa and 

 Micranthes of Tausch, giving the name Arabidia to the combination. Like 

 his predecessors, however, Ledebour based his classification wholly on char- 

 acters of minor importance, which, together with the combination he intro- 

 duced, added still more to the confusion in the classification. 



This first attempt at combination thus initiated by Ledebour reached 

 its culmination in the work of Engler, who combined into a single group 

 the sections that had been separated for nearly fifty years. 



RosENDAHL (25) Considered the sections Micranthes, Arabidia, Spatu- 

 laria, and in part Hydatica, as subsections of Engler's section Boraphila. 

 He regarded the structure of the floral axis and of the carpels as the 

 principal diagnostic characters of Engler's section. 



Contemporaneously with Rosendahl, S.mall (27) published a revision 

 of the group, in which he digressed from the arrangement of previous 

 writers by restoring the genera Micranthes and Spatularia of Haworth, and 

 at the same time adopting Rafinesque's (24) genus Heterisia for the 

 monotypic Heterisia (Saxifraga) Mertensiana Bongard. Small here rec- 

 ognized the hypogynous character of the species included in these two 



