r 



THE IDENTIFICATION OF WALTERS GRASSES. 



37 



about as in the preceding, slightly hairy. Leaves about 

 the same shape, sparsely hirsute. Sheaths quite hirsute 



margin 



P 



vosurn 



P. commutation Schultes) 



m ay 



be, as this is common in that region. P. brevi folium L. 

 is from India, but Linnaeus' description is copied by 

 Walter. 



nudum 10. culmo geniculato sesquipedali erecto, floribus 



solitariis remotis verticillatis, pedunculis 

 longissimis ; foliis caulinis nullis. 



No specimen. Muhlenberg suggests that this may be 



Schult 



P 



Bosc), while Pursh refers it doubtfully to P. meli- 

 carium Michx. It may be P. cognatum Schultes, as the 

 culm is about a foot and a half hiffh with the flowers 

 (spikelets) on long peduncles, and the leaves are mostly 

 near the base of the culm. I do not know just what is 

 P. melicarium Michx. From the description it might be 

 P. Mans Ell. Walter's plant can scarcely be this, as 

 suggested by Pursh, for the description of P. melicarium 

 states that there is a rudimentary flower at the base of 

 the perfect flower, "and M ramulis raris, subappressis." 



coloratum 

 11. 





fol. amaris, longissimis, fauce villosis; culmo 

 folioso, 3-pedali, panicula patentissima 

 laxiuscula ramis filifonnibus. Antheris 

 croceis; pistillis saturate violaceis- 



The 



No specimen. Referred by Pursh to P. virgatu\ 

 and by Michaux doubtfully to the same species, 

 statement as to the bitter leaves applies to P. amarum 

 Ell., but the remainder of the description applies also to 



P 



Elliott states that Walter has confused 



the two species which the former now separates. P 



