120 Chase — Notes on Genera of Panice&. TV. 



"appears to have been the type of the proposed genera Acicarpa, Eaddi, 

 Eriachne, Philippi, and Holosetum and Mesosetum Steud." (It is in fact 

 the type of only the first-named of these. ) In the Genera Plantarum 

 (Benth. & Hook. Gen. PI. 3 : 1101. 1883) he says that Mesosetum is per- 

 haps (forte ) referable to the section Trichachne of Panicurn. At the same 

 time Bentham (Linn. Soc. Journ. Bot. lit : 42. 1881) establishes as a 

 section of Panicurn his Diplaria comprising " P. rottboellioides, H. B. K., 

 P. exaratum and P. fcrrugineum Trim, P. pappophorum, Nees, and a 

 few others." In the Genera Plantarum (1. c. ) this section is described 

 and the same species mentioned as belonging to it. Dalla Torre and 

 Harms (Gen. Siphonog. 14. IflOO) also include Mesosetum, together with 

 Alloteropsis Presl, Coridochloa Nees, Eriachne Phil., as well as Acicarpa 

 Raddi and Trichachne Nees which properly belong there, under Pdnicum, 

 section Trichachne. Steudel himself failed to see the identity of his 

 Mesosetum cayennense with P. rottboellioides or its affinity to the other 

 species of this group which he includes under the section Harpostachys 

 (Syn. PI. Glum. 1 : 55. 1854). It was only Prof . Hitchcock's examination 

 of the type specimens of Mesosetum cayennense and P. rottboellioides, both 

 now in the Herbarium of the Museum d' Histoire at Paris, that revealed 

 the specific identity of the two. It is unfortunate that SteudePs generic 

 name, with its inadequate description, must stand for this well-marked 

 genus. But if the names of all incorrectly described genera were rejected 

 the nomenclature of the Gramineae would undergo many changes, for 

 this family seems particularly to have suffered from the misunderstanding 

 by authors, of the morphology of the parts of its inflorescence. Anthaen- 

 antia, Ichnanthus, Alloteropsis and Pentarrhaphis are examples of valid 

 genera incorrectly described by the authors who bestowed the names we 

 use for them. 



Description. — Inflorescence a single, terminal, erect, spike-like raceme, 

 the spikelets subsessile, solitary, alternate in two rows on one side of 

 a three-angled, rarely winged, tortuous or zigzag rachis, the spikelets 

 with the back of the fruit turned from the median line of the rachis 

 (the first glume towards it), ventricose on the side toward the rachis, and 

 fitting into its concavities, the back of the spikelet flat or nearly so; 

 glumes 3- to 5-nerved (when 5-nerved the lateral pairs of nerves approxi- 

 mate) the lateral nerves often uniting with the midnerve below the acute 

 or acuminate summit, one or both usually clothed on the margin with 

 stiff hairs; sterile lemma like the glumes in texture and the distribution 

 of the pubescence, usually appearing L'-keeled from the strong lateral 

 nerves and firm lateral internerves and thin or hyaline middle internerves 

 and weak midnerve; sterile palea wanting except in the section Bifaria ; 

 fruit pointed, ventricose on the face (palea side) usually straight on the 

 back, the lemma and palea less indurated than in Panicurn, the flat 

 margins of the lemma not hyaline. Slender perennials with narrow leaves, 

 the uppermost reduced to a bladeless or nearly bladeless sheath. 



This genus, like Brachiaria, differs from Panicurn in the strictly race- 

 mose inflorescence and reversed position of the spikelets, and further in 

 the form of the spikelets, swollen or ventricose on the face (or first glume 



