PRUNUS RIVULARIS 207 



a form of Prunus Americana" The whole group has also been 

 roughly referred to /'. gkmdvlosa, Torr. & Gray, bul this is evi- 

 dently a mistake. Mr. T. V. Sfunson has given this quesl 

 Berious study, and has concluded that all these varieties are 

 derived from /'. rmtlaris, Beheele. This is a somewhal start- 

 ling decision, and extremely important if true. The facts are, 

 however, first, that we are yet too poorly acquainted with this 

 species to make critical comparisons ; secondly, that 8che< 

 description, made at a ad hand from Lindheimer's speci- 

 mens, is do! sufficiently precise to preclude mistakes ; thirdly, 

 that the description,* what there is ,,f it. fails, in important 

 particulars, t.. tit the varieties in question : and fourthly, that 

 many of these varieties have originated in localities where it 

 is almost impossible t.. believe that /•. rwularis could be grow- 

 ing, s.-,. pages 223, 224. 



To particularize further, the National Berbariumt contains 

 only the following specimens : Those of Lindheimer, collected 

 in western T.-xas i„ 1846; one by Hall, from Dallas; two by 

 Wolf, collected in Illinois in 1875, and very possibly cultivated 

 specimens : and one of doubtful authenticity, by Thon 

 ler, from Manhattan. Kansas. Other herbaria Beem t.. have 

 n.. better representation of th< and this could hardly 



I..- the --a-' were it s.. common and s.. widely distributed as 

 furnish the well-known cultivated varieties mentioned above. 



•sin.',, this description i- inaccessible to many students, it will be ml] to 

 ■ I be it here : 



'''""•" >■•"■ di«i rami toga- 



_-.ii.ri niti.ii cinerei verruculosi, ret-ruca parrs wmfertss. Petioli glandalosl 



basl 

 «'•'"" ipra glabra, 



nmas Borifene aphj . i-. 



mi.™ globoea glabra nit 



'•ehafiUch , ner aber Jedesmal In Ifenge raaam- 



tlen »uf H&geln. - ,_, 



■ 

 I mil honig 



imlned foi • 



Mi l.-.-r. ,|| |.. 



