1916] Fernald & Weatherby — Puccinellia 3 



In attempting to determine the large collections of Puccinellia 

 which have accumulated at the Gray Herbarium, especially from the 

 shores of the Gulf of St. Lawrence, the writers have found it impossible 

 to place many of the plants with the species commonly recognized 

 as occurring in eastern America. It has consequently been necessary, 

 in order to make the identifications of these plants as certain as pos- 

 sible, to study with some thoroughness all the species to which they 

 are nearly related or with which they may be conspecific. This 

 study has been greatly facilitated by the loan, through Dr. J. M. 

 Macoun, of the large northern collections of the Geological Survey of 

 Canada, by Dr. M. O. Malte of his private collections, and, through 

 Mr. Bayard Long, of the representation in the herbarium of the 

 Academy of Natural Sciences of Philadelphia; and we here express our 

 appreciation of the opportunity to examine this most important 

 material which, with the collections of the Gray Herbarium and of 

 the New England Botanical Club, has been the chief basis of our work. 



It quickly became apparent that much of the difficulty heretofore 

 experienced in efforts to differentiate the species of Puccinellia has 

 arisen through attempts to divide the plants too generally upon the 

 habit of the inflorescence. Many species which in full maturity have 

 the branches of the panicle widely divergent, in the young condition 

 have close inflorescences with ascending or even appressed branches; 

 so that attempts to use this character often lead to confusion. In 

 searching for more fundamental characters which should be constant 

 in all the material of seemingly identical plants many fruitless experi- 

 ments were made; but finally the results resolved themselves into a 

 series of differential points which seem to be reasonably definite and 

 to divide the complex of material upon natural and geographically 

 consistent lines. These characters have been tested through a pro- 

 longed study, which has occupied portions of three years, and although 

 here indicated chiefly for the species which occur in eastern North 

 America (south of Hudson Straits) they will prove important, we are 

 sure, in the differentiation of the species of northwestern North 

 America, Eurasia and the Arctic, in each of which areas there are 

 several species not here discussed. 



In the region specially covered by this paper (the area south of 

 Hudson Straits) there are eleven well defined species of Puccinellia, 

 each, as already implied, occupying a consistent geographic area. 

 In two species, P. maritima and P. phriiqanodcs, the anther is 1.5- 



