41 



Culms very leafy, sheaths usxially iiim-h overlapping and furnuhed iiith a rather con- 

 spicuous pilose-pubescent ring (tt tlie summit; blades rather broadly auricled at the 

 base. Spikelets and flowering glumes rather broader than in the species. The 

 pubescence at the base of the flowering glume is slightly denser than elsewhere. 

 In other respects like the species. 



Type No. 222, collected l)y L. H. Pammel, Dakota City, Iowa. 



General di8tril)ution about the same as for the species, liut apparently reaching its 

 greatest development in numbers west of the Mi-ssissippi. 



Specimens examined. — Connecticut: South Glastonbury (Frances Wilson 124). Nei.v 

 YorJ:: Oxford (F. V. Coville). Pennsylvania: Easton (T. C. Porter). Minnesota: 

 Forest :Mills (C.A.Ballard). South Dakota: Redfield (David Griffiths 74); Big 

 Stone (^Ir. & Mrs. T. A. Williams; James River, Brown County (David Griffiths 

 784); Canning (T.A.Williams). Montana: Smith River (F. Lamson-Scribner 

 77). Nebraska: near Thedford (P. A. Rydberg 1775 — this number is the one on 

 whicli Rydlierg founded the new combination cited above. In character of 

 spikelets and in the rather narrow'er leaves than usual it approaches B. porteri, 

 but in other respects it is this variety; Holt County (F. Clenaents 2824). Iowa: 

 Forest City (B. Shimek 62) ; Ames (C. R. Ball 4) ; Jackson County (B. Shimek 34) ; 

 Dakota City (L. II. Pammel 222). Missouri: (H. Eggert); Jefferson County 

 ( H. Eggert 606) ; Allenton (G. W. Letterman 14). 



This form connects the species with B. richardsoni and B. porteri, the more numerous 

 leaves, the overlajiping sheaths, and larger panicle separating it from the latter. 

 The overlapping sheaths and more equally distributed pubescence of the flower- 

 ing glume distinguishes it from the former. The citation of B. altissiinus Pursh 

 as a synonym is based upon the examination of specimens in the herbarium of 

 the Philadelphia Academy marked "Herb. Pursh." The specimens agree per- 

 fectly with Pursh's description and are apparently authentic. 



21/1 BROMUS PURGANS ? TEXENSIS n. var. 



A slender i)lant, somewhat geniculate at the base and 3-4 dm. high. Sheaths short 

 pilose-pubescent; ligule 1-1.5 mm. long, laciniate; blades linear, pubescent 

 throughout, about 6-10 cm. long. Panicle small, somewhat nodding, with only 

 3-5 spikelets in the specimen at hand. Spikelets terete at first, then lanceolate 

 and somewhat laterally contracted; empty glumes acuminate, scabrous, the 

 lower 1 -nerved, 7 mm. long, the upper inequilateral and mucronate at the apex, 

 9-10 mm. long; flowering glume 7-nerved, 10 nun. long, sparsely and coarsely 

 scabrous, very short dentate at the apex ; awn straight, 6-7 mm. long ; palea equal- 

 ing its glume, keels finely serrate. 



Type No. 230, collected by G. Jermy, Bexar County, Texas. 



This is the only specimen of this form we have seen. More material is necessary to 

 determine definitely its status and relationships. It may prove a good species. 



21 r. BROMUS PURGANS INC ANUS n. var. 



This is very near B. purgans latiglumis, differing from it in having the sheaths 

 densely soft pilose-pubescent. It also jmssesinto B. porteri lanatipes in the south- 

 west. It is generally separated from that by its broader leaves and narrower 

 empty glumes, Avhich are like the flowering glumes sparsely pubescent. 



Type No. 3, collected by J. Wolf, Canton, Illinois. 



General distribution apparently about the same as for the species. 



Specimens examined. — Pennsylvania: Easton (T. C. Porter). District of Columbia: 

 High Island (F. L. Scribner). Ohio: Pittsfield (A. E. Ricksecker). Iowa: Fay- 

 etteCounty (B. Fink, 414). South Dakota.: Union County (E. J.Wallace). Te.ms: 

 Chenates (G. C. Nealley), poor and not typical. 



