1884.] NATURAL SCIENCES OP PHILADELPHIA. 289 



OBSERVATIONS ON THE GENUS CINNA, WITH DESCRIPTION OF A 

 NEW SPECIES. 



BY F. LAMSON SCRIBNER. 



Two species of Cinna,^ common to the northern regions of 

 both the old and the new world, have long been recognized. They 

 are G. arundinacea L. and C. pendula Trin. The latter, the 

 more common of the two, has been reduced to a variet}^ of the 

 first-named, by some authors; but, aside from a marked diversity 

 in habit, there are important differences in the characters of the 

 spikelets quite sufficient to warrant a specific distinction. In 

 G. arundinacea the spikelets are larger, firmer in texture, more 

 strongly scabrous, more prominently nerved and there is a 

 decided inequality in the outer or empty glumes, while in G. 

 pendula these glumes are equal or nearl}' so. In both the floret 

 is stipitate, or raised on a short stallc above the insertion of the 

 empty glumes ; in other words, there is a slight elongation of the 

 axis of the spikelet between the two empty glumes and the flower- 

 ing glume. Mr. Bentham, in his " Notes on Gramineaj," states 

 that in G. arundinacea there is frequentl}^ a continuation of the 

 rhachilla in the form of a short naked pedicel behind the palea : 

 a character, he adds, that he has never seen in G. pendula. In 

 my own studies, I have found this prolongation of the rhachilla, 

 a common, not constant, character in both species, I have 

 observed it in the spikelets of G. pendula from Maine and from 

 Oregon ; in fact, my own observations would lead me to say that 

 it appears more frequently in that species than in G. arun- 

 dinacea. 



In regard to the variations in these two species something may 

 be said. G. arundinacea exhibits considerable diversity in the 

 size and diffuseness of its panicle and the spikelets, which range 

 from a little over two to nearly three lines in length, vary in 

 color from pale green to dark purple, but those characters men- 

 tioned above as distinguishing this species from G. pendula, 

 remain constant. There is greater variation of the panicle in 

 G. pendula, and also in the size of the spikelets ; these, however, 

 never reach two lines in length, and, although the empty glumes 



' Cinna macroura Kth. and other grasses that have been placed in this 

 genus, are now referred to Epicampes or Deyetixia. 



