54 Bulletin VII. 1. 



lished by Lamarck, and I venture to insert Michaux's description 

 as it affords some additional characters: Plant glabrous, with slen- 

 der, erect, simple culms. Sheaths barbate at the throat. Leaves 

 very few, remote linear-lanceolate. Panicle capillary, somewhat 

 contracted, much branched, glabrous. Spikelets very small, ovate- 

 obtuse, very minutely puberulous, outermost glume scarcely per- 

 ceptible.* 



Note. — Here we have a species based upon an early stage of the plant 

 while the stems are yet unbranched. The original descriptions of P. pu- 

 bescens^ nodifloriim^ ramulosnm, barbulatum^ dichotomum^ etc., were drawn from 

 more mature specimens after the culms had become much branched, and 

 the difficulty in determining these plants has largely arisen from this 

 fact. An exact knowledge of the species in this group can hardly be ac- 

 quired excepting through observation in the field at different periods of 

 their development. I have found little variation in the several species 

 arising froin soil or exposure, but the differences between the early growth 

 and that which appears later is often so great that one unacquainted with 

 the plants in the field would never suspect the identity of the two forms. 

 In all, the panicles which terminate the branches are fewer-flowered and 

 more simple than that of the primary stem. In P. latifolium and in P. 

 commutatum the leaves of the branches are sometimes broader than those of 

 the primary stem, but in the others these leaves are usually much smaller 

 in every way— in all, these leaves are more crowded. 



30. Panicum dichotomum Linn. 



Panicle simple, culms dichotomously branched (Linn. Sp. PL I., 

 p. 58). Linnaeus adds that this grass resembles a little tree, the 

 stem being simple below and fasciculate-branched above. This is 

 all we have for the identification of this species among the many 

 forms in this group, in which, as already stated, the culms become 

 much branched as the season advances, the primary panicles dis- 

 appear and the later panicles on the branches are always simple 

 and few-flowered. The remark that the culms are simple below, 

 giving to the grass the appearance of a little tree, affords the best 

 clue to the plant referred to by Linnaeus. As I understand the 

 species, it may be characterized from my material as follows: Gla- 

 brous throughout. Culms slender, becoming almost wiry, simple 

 below, becoming much-branched above, branches spreading. Leaves 

 thin, pale green, two to four inches long, those of the branches 

 smaller, sometimes nearly filiform, and widely spreading. Panicle 

 diffuse, branches naked below, lateral panicles simple and few- 

 flowered. Spikelets scarcely a line long; ovate-oblong, outer 

 glumes glabrous. P. barbulatum Michaux may be only a variety 

 of this species, with the nodes bearded. 



31. Panicum barbulatum Michx. 



Culms one to three feet high, slender, finally much-branched, 

 smooth ; nodes densely bearded with defiexed hairs ; sheaths 

 sparingly pilose or smooth, much shorter than the internodes; 

 ligule a fringe of short hairs, hardly a line long; leaf-blades lanceo- 

 late, two to five inches long, three to six lines wide, spreading, acute, 

 narrowed at the base, glabrous; margins minutely serrulate and 



*In tbe case of Panicum nitidum and the three following species, It has appeared best to 

 present the descriptions of the original authors. The limitations of these species are 

 now exceedingly doubtful, and It Is hardly possible to Identify them without consulting 

 the types. 



