16 MEMOIRS FROM THE DEPARTMENT OF BOTANY OF COLUMBIA UNIVERSITY 



from such species as P. sterilis and P. ovalis only in the receptacle, which becomes fleshy 

 in fruit. The characters that distinguish Duchesnea from Potentilla are still less marked, 

 viz., the enlarged and red-colored receptacle. Since in Comarum, which is included in 

 Potentilla by these authors, the receptacle also enlarges, the only remaining character 

 would be the red coloring. 



Professor Greene^ when carrying out in practice what was suggested by Bentham & 

 Hooker and Baillon, and renaming the known species and describing several new ones, 

 in his introduction pointed out the great variety in habit in the genus Potentilla, using 

 P. Anserina, P. fruticosa and P. arguta as good illustrations. He also showed that Hor- 

 kelia, Sibbaldia and Ivesia did not differ more in habit from Potentilla than the species 

 mentioned differed from each other and from others of the genus. He was perfectly 

 right in this view. For my part I think that those three species differ in essential char- 

 acters more from each other and from Potentilla proper, than do Sibbaldia, Horlcelia 

 and Ivesia, especially the latter. I should regard them as the types of three good genera. 

 Among other characters separating them from Potentilla is a different position of the 

 style, a character which even in Rosaceae is taken as the principal distinction be- 

 tween the tribes Pkuneae and Chrysobaxaneae. ^ 



Only two courses can logically be taken. Either the whole tribe, Fragaria also 

 included, must constitute a single genus, or else both Potentilla and Ivesia be divided into 

 several genera. I have taken the latter course, for the following reasons : 



1. Nature was not made to order so as to fit a special human system, and the Hues 

 between genera or species are seldom as well marked in nature as the scientist wishes 

 they were, or even as he draws them. Nearly everywhere in nature there are found in- 

 termediate forms even between genera. The larger these are and the more characters 

 are used to distinguish them, the more such intermediate forms there will be found. 

 There is, therefore, a tendency among both botanists and zoologists to acknowledge 

 many and small genera, based on few, or even single characters. Each genus will then 

 consist of only such forms as are very closely related and the system will gain in clear- 

 ness. 



2. Potentilla has already too many species to be handled conveniently and still 

 more are discovered every year. It would, therefore, be well to divide it simply for 

 convenience. 



3. The genus Potentilla, in its broader sense, contains plants of so many different 

 types that it can scarcely be regarded as a systematic unit of the rank that is called a 

 genus in general or as analogous to the genera Fragaria, Rosa, Rubus, and Dalibarda. 



'Pittonia, 1 . 97-106. ^ jjjgge tribes are now regarded by many as good families. 



