. 97 
the number of florets the cases quoted in Bentham and Hooker’ 8 
Genera Plantarum (S. compressus and S. serotina) and in 
Martius’s Flora Brasilliensis (S. ramosissima) concern sa 
individual spikelets in otherwise oe PS ators and seem 
to be altogether rare. I myself have seen 2-flowered spies 
only in a specimen of S. compressus which’ was attacked by some 
fungal disease. Instances of a similar exceptional addition of 
florets occur in Pani ‘tins Andropogon and Oryza, genera with 
an otherwise perfectly stereotyped number of Rosets: and they 
have never been considered to affect the character of those genera. 
However close the spikelets of the Derendingen grass may 
approach those of Sporobolus in the structure of the parts, they 
contrast so decidedly in the number of florets that to include the 
grass in Sporobolus would vitiate our well established conception 
of that genus. This contrast is moreover enhance h 
grouping of the spikelets in pairs on the somewhat distant spike- 
like and themselves spicate branches of the inflorescence, a 
disposition not known in Sporobolus. Thus the Derendingen 
grass appears to represent a type concording in many ways with 
Sporobolus, yet clearly and discontinuously detached from it in 
other respects, and this condition will be best expressed by con- 
ceding generic rank to the new type and placing it next to 
Sporobolus. The fact that it grew among a considerable number 
of aliens of lama origin renders it highly probable that this 
grass too came from the same source. The following is a 
technical descrtsittine of the new genus and of its only species. 
The genus has been named Thellungia in acknowledgment of the 
excellent work Dr. ae has done in connection with the 
adventitious flora of Euro 
Thellungia, Stap} |Gramineae-Sporoboleae } Ben nov.; affine 
ee sed spiculis 3-4-floris bene distinctum 
ulae lateraliter se Sli , per parila in ramis brevibus 
oedae ge spicam compositam referentis subsessiles, imbri- 
catae, continuae cum pedicellis ; rhachilla tarde disarticulata supra 
glumas et inter anthoecia, glabra, pee? ultra basin anthoecii 
summi producta. Anthoecia 3-4, omnia g. Glumae 2, delicate 
membranaceae, subaequales, 1 -nerves, fate carinatae. Valvae 
elumis simillimae, nisi superiores breviores.  Valvulae quam 
valvae conspicue breviores, tenuiter 2-carinatae, inter carinas 
subplicatae. Lodiculae 2, late cuneatae, carnosae.’ Stamina 3. 
Ovarium glabrum ; styli perbreves, terminales ; stigmata plumosa, 
brevia. -Caryopsis libera, nuda cadens ut videtur; pericarpium 
tenue, siccum semini adpressum, humefactum expansum, semen 
more utriculi circeumdans. Semen a latere compressum, sectione 
transversa cuneatum; hilum punctiforme, basale. Embryo ad 
e 
i men attingens.—Gramen gracile, perenne; foliorum 
lJaminae angustae, vernatione convolutae ; ligulae a marginem 
ciliolatum redactae. Infloreseentia angustissima ; rami solitaria, 
e basi floriferi, axi communi subadpressi et eius internodiis paulo 
longiores vel superiores breviores et magis appressi. iculae 
nitentes, mediocres 
Species unica, verosimiliter coceslionie: 
