234 ON CALAMAGROSTIS LANGSD011FF1I. 
whatever of an upper floret ; and not only do the flowers precisely 
accord with Andersson's analytical figures of C. phragmitoides (Gram. 
Scandin. t. ix. f. 97), but they are absolutely undistinguishable under 
the microscope from those of the plant distributed under that name 
(FL Lapp. Exs. n. 279) by Professor Andersson himself, to whom I 
am indebted for a specimen, and the two plants agree perfectly in all 
respects. Hence this species must be added to the flora of North- 
Eastern Asia. But the further question arises whether these two re- 
puted species are really distinct. It seems to have been overlooked 
that Trinius originally described his C. purpurea as "processu nullo," 
and C. Langsdorjjii as "processu longe barbato" (Gram. Uni- et Ses- 
quifl. 225, 229), although Grisebach (1. c.) attributes to each a u mo- 
mentum minutum, pilis elongatis terimnatum." Fries says (Summa 
Veg. Scand. 239), "Notae vulgo autem, etiam ad sectiones discernendas 
citatse (v. c. situs aristae, rudimentum secundi floris) passim occiden- 
tales;" and Andersson {op. cit. 80), "Friesius aperte demonstravit 
praesentiam v. defectum flosculi secundi in hoc genere minimi esse 
momenti." Moreover, Maximovvicz has described (Prim. FL Amur. 
324) a viviparous variety of C. purpurea, destitute of a rudimentary 
floret. It would be verv desirable to ascertain whether the presence 
or absence of this organ is really of even specific value in this intricate 
and protean genus. Unless the nomenclature adopted in labelling 
Drs. Hooker and Thomson's ' Indian Grasses,' and that employed in the 
enumeration of those of the Linnean herbarium, are to be taken as 
proofs of Colonel Munro's matured views, I know of no eminent living 
agrostographer who now regards Deyeuxia as a genus, except Parla- 
tore, who still distinguishes it (FL Ital. i. 211) by the rudimentary 
upper floret and geniculate awn. But both Fries and Andersson deny 
that the two characters constantly coexist ; and, in fact, the strnight- 
awned C. lanceolata, Roth, is recorded as having sometimes a rudi- 
mentary flower. C. stricta, Spr., has a straight awn and an upper 
floret ; whilst in C. Lapponica, Wahl., the awn is, as it were, interme- 
diate between the two sections. Dr. Hooker has recently (Handh. x . 
Zeal. Fl. 329) reduced the species of the southern hemisphere to 
Agrostis, a view which had been previously taken by Trinius in I» s 
* Agrostidea.' Of these, some, as A. Billardkri and A. amide, have 
a twisted awn ; others, as A. C/utmisso)iisi\\id A. setifolia, a rigid one. 
The truth is, that the genera of Agrostidea have been unreusonab.v 
