Chase — Notes on Genera of Panicece. IV. 113 



petrosum Trin. and the other two species here listed. Judging from 

 Kunth's description and plate 39, and allowing for evident error in each, 

 Thrasya paspaloides is closely related to P. thrasyoides Trin. Kunth 

 places his Thrasya as the last genus of " Sectio I. Paniceae" (ahout the 

 equivalent of our subfamily Panicoideae) immediately following M anisuris 

 granulans (Hachiocldoa). He diagnoses Thrasya as being 2-flowered, 

 having two glumes, "superior profunde bipartita," the halves aristate 

 below the apex, " inferior [gluma] integra mutica." The male floret is 

 said to have but a single palea [both valves of a floret being termed paleae] . 

 From dissections of spikelets of P. thrasyoides and from plate 39 the follow- 

 ing conclusions are reached: Knnth overlooked the small, hyaline first 

 glume (which in P. thrasyoides is more or less buried in the cleft of the 

 sterile lemma and might easily escape notice); his entire, awnless, lower 

 glume is the second glume; his deeply divided upper glume, the sterile 

 lemma, the aristae of the halves being not awns (prolongations of fibro- 

 vascular bundles) hut stiff, cpiill-like hairs; the single palea of the male 

 floret is the sterile palea. It is strange that Kunth, failing to note the 

 first glume, did not, nevertheless, count it as obsolete, since failing to do 

 so, and counting the second glume as the first, he describes a spikelet in 

 which the scales are not distichous, for his male floret is immediately 

 above the (supposed) second glume instead of on the opposite side above 

 the first. The margins of the sterile palea (in the allied P. thrasyoides) 

 are so narrow and so readily torn from the hyaline middle portion that 

 the fact that they turned toward the supposed second glume might escape 

 observation. Kunth describes this " flos masculus " as " subtrinervia " — 

 being a palea it has no midnerve, where Kunth evidently looked for one. 

 But even if this [supposed] incongruous structure of the spikelet escaped 

 him (if it did escape him) Kunth considered his Thrasya a most curious 

 grass. He states that though it resembles Paspalum platycaule in habit, 

 the structure of the flowers is so singular as to he widely different from 

 that of all known genera. Nees (Agrost. Bras. 93. 1829) in his diagnosis 

 of Thrasya notes a minute, scale-like inferior glume " (a cl. Kunth 

 neglecta.)" The further important character that Kunth neglected, that 

 is the alternately reversed position of the spikelets on the axis, Xees takes 

 note of in his specific description of Thrasya hirsuta (based on Panicum 

 thrasyoides Trin. ), though Trinius failed to do so. 



Description. — Inflorescence a single, terminal, more or less arcuate, 

 spike-like raceme ; spikelets apparently suhsessile and solitary, in a single 

 row on one side of a channeled, more or less winged rachis (the wings 

 partially embracing the row of spikelets), hut actually in pairs (the 

 spikelets of each pair back to back) the pedicel of the primary spikelet 

 adnate to the midnerve of the rachis* (the spikelets spreading from the 



* To account for the alternation in position of spikelets. this was the interpretation 

 arrived at by the writer from an examination of P. thrasyoides and P. petrosum, but I 

 should not have ventured to give it at this time had I not found that it is the conclusion 

 of Prof, Hackel (Oesterr. Bot. Zeitschr. 51 : 368. 1901) in the case of Panicum campylo- 

 stachyum, in which species the paired arrangement is more evident than in P. thrasyoides 

 and P. petrosum. 



