FEB. 1903. PLANTS YUCATAN^; MILLSPAUGH & CHASE. 



and internerves pubescent with appressed hairs; third 3.4 mm., 

 7-nerved, the lateral ones often indistinct, margins and second inter- 

 nerves on each side pubescent with 

 short appressed hairs or clothed 

 with long silky hairs or sometimes 

 almost glabrate, first and second 

 nerves glabrous, scabrous or 

 pubescent; floral glume nearly 

 equal to third, pale brown with 

 white margins, papillo-striate, acu- 

 minate; palea equal and of like 

 texture. Grain translucent pink- 

 ish white, ovate-ellipsoid, rounded 

 at apex and base, .8 x 2. i mm. ; in 

 section convexo-concave. 



Hab. Merida, Aug. 27, 1865, 

 Schott 609 (Panicum marginatum 

 Link non R. Br. non Vahl, Field Col. 

 Mus. Bot. 1:353), Aug. 20, $86, 

 ruins of Uxmal, Sept. 16, 1865, 

 7J5 {Panicum sanguinale L. ibid.}; 



Yucatan loc. ignot. Gautner 1029 (Panicum sanguinale L. ibid.}, Izamal, 

 1034, Chichankanab 1292, San Anselmo 2143; rock-strewn arid 

 plain south of Progreso, Millspaugh PL Uto. 1700. 



An extremely variable species with apparently no constant char- 

 acter except the fruiting glumes and the grain. After examination of 

 a large amount of material from North and Central America and the 

 West Indies, the forms known as Digitaria marginata Link and fim- 

 briata Link are here included under Syntherisma sanguinale. (Kunth, 

 Enum. 1:82, makes D. marginata Link a synonym of Panicum sangui- 

 nale]^., but D. fimbriata Link he holds distinct.) Individual speci- 

 mens can be selected showing apparently distinct species, but others 

 show such complete intergrading that no combination of characters 

 has been found to hold. No. 1029 Gaunter, from which the illustra- 

 tion is made, shows the two forms of spikelet, fimbriate and nearly 

 glabrate, borne on different panicles from the same branching culm. 

 The other distinctions: length of second glume, scabrous or smooth 

 nerves, and size of spikelet, have not been found to hold good; spike- 

 lets of sanguinalis equaling those of fimbriata, excluding the long 

 hairs of the latter. The first and second nerves of third glume are 

 found to be scabrous usually on northern plants, but pass into the 

 softly pubescent nerves of the southern form which is not "fimbri- 

 ate"; where the internerves are pubescent the nerves are usually 

 glabrous. No. 19068 Herb. Field Col. Mus. Coll. C. F. Millspaugh, 

 Morgantown, W. Va. , July -30, 1890, shows spikelets with scabrous 

 nerves of third glume and others with glabrous nerves and pubescent 

 internerves (species sanguinalis and fimbriata) on the same rachis, and 

 in one case the pair of spikelets shows one of each form. Taken 

 altogether, however, the specimens examined show that the "fimbri- 

 ata'" form usually has a longer axis of inflorescence, panicled rather 

 than digitate, with spreading- racemes, while the typical sanguinalis 



