176 The Ottawa Naturalist. [November 



Nevada, Siberia, Norway and Greenland." Two varieties, 

 grandis and Iwlopetala^ are said to occur in S. E. Arizona, Cali- 

 fornia and Oregon, and since Pallas' G. ciliata'^ is by Gray con- 

 sidered as identical with G. serrata, the species should occur also 

 in the mountains of Caucasus. 



When high northern or arctic plants are found farther south, 

 they are as a rule confined to very high mountains and above the 

 timber-line, but if the plant described by Gray were the true G. 

 serrata it would be equally abundant much farther south and at 

 low elevations, in for instance, Canada and the United States. 

 It was not, however, this incredible geographical distribution 

 alone that made the writer suspicious in regard to the identity of 

 the American so-called G. serrnta with that of Europe, but also 

 the fact that our material, which we some years ago collected in 

 the Rocky Mountains of Colorado, differed so very much from our 

 European specimens. The difference appeared to us so striking 

 that the que>tion of considering the Rocky Mountain plant as a 

 meie geographical variety was at once excluded. This view be- 

 came especially strengthened when we learned from our friend, 

 Professor Wille in Christiania, that none of our specimens could in 

 any way be considered as belonging to the species '■'■serrata " as 

 this occurs in northern and arctic Norway. By studying the 

 several specimens of the true G. serrata, which Professor Wille 

 kindly sent to the writer, we felt convinced that the diagnosis in 

 the Synoptical Flora was based on specimens from America of 

 some allied species without being compared with the European 

 plant. In order, however, to make our investigation as careful as 

 possible, the writer applied to the Geological Survey Department 

 at Ottawa for the loan of the entire collection of this group, as 

 represented in the British provinces, which was courteously 

 granted through the kindness of Mr. James M. Macoun. Being 

 unable to find any trace of G. serrata in this comprehensive collec- 

 tion we felt obliged to consult, also, the herbarium of Gray and 

 of the United States' National Museum, of which the complete 

 material of the American G. serrata was kindly placed at our 

 disposal. 



3 Pallas, Peter Simon, Flora Rossica. St. Petersburg, 1784. Vol. i, 

 p. 101, plate 92, fig, 2, 



