Holm, Some new anatomical chnracters for certain Gramineae. 123 



"beneatli the bulliform cells in the fiirrows. — The mesophyll 

 consists of palisades, which radiale towards the center of the 

 mestome-bundles in all the species, which we have examined. 



A Single and green parenehyma-sheath is observable around 

 the mestome-bundles; it is generally large-celled, but the cell- 

 walls are either thin or distinctij thiekened, the latter being 

 characteristic of S. airoidesy hrevlfolius , cuspidatus, pungens and 

 vaginaeflorus ; a parenchyma- sheath of small and thin walled cells 

 was observed only in ^S. junceus. 



Inside the parenehyma-sheath is a mestome sheath, noticed 

 in all the species, *) of which the inner cell-walls are usually 

 thiekened, very heavily in S. airoides^ asper, indicus, junceus and 

 pungens. — As to the location of the mestome-bundles, they are 

 in most of the species confined to the ribs, but may also be seen 

 in the furrows in a few species {S. airoides, asper and vaginae- 

 florus) ; a transverse section shows their outline to be oval or 

 orbicular, the latter form being the most frequent, but often mixed 

 with the former in the same leaf [S. asper, hrevifolius, cuspidatus, 

 junceus etc.). The mestome-bundles are thus either all orbicular, 

 or this form may occur together with the oval, but we observed 

 no instance where the latter form was the only one represented. 



Calamovilfa longifolia (Hook.) Scribn. (Plains, Denver, 

 Colorado). 



This genus is also a representative of the ^Euagrosteae"' and 

 is very nearly related to Ammophila Host. {Psamma Beauv.) by 

 the chartaceous flowering glume ; but it has the large and loose 

 panicle of Calamagrostis, to which the species was also refered 

 by Hook er. — There is still another species of this little genus, 

 which by Gray was placed with the former as a section of 

 Calamagrostis, but Ha ekel proposed the segregation of these 

 species to form a genus ^^Calamovilfa": „Empty glumes unequal; 

 flowering glume one-nerved: rhachilla not prolonged". We have 

 only examined C. longifolia, which grew socially on the plains 



*) In describing the leaf-structure of Sporobolus Virginicus (L.) Ktli, 

 Professor Warminp (Halofyt-Studier, p. 227) states that the mestome- 

 bundles, viewed in transverse section, are arranged in a zig-zagged line, 

 the larger being located in the projecting ribs, the smaller in the furrows; 

 that the cells of the mesophyll do not present any very pronounced shape 

 as palisades, and that epidermis of the lower face is generally very large- 

 celled. He compares this structure with that of S. spicatus, figured by 

 Volkens (1. c). But the specimens which we examined of S. virginicus 

 from Florida exhibit a structure so dilFerent from that described by Professor 

 Warming, that the material examined by him must have been wrongly 

 identified as of S. virginicus. The leaf of real S. virginicus has no mestome- 

 bundles in the furrows. The mesophyll represents a true and very distinct 

 palisade-tissue, the cells of which radiate towards the center of each mestome- 

 bundle, besides that the cells of the epidermis are small on the lower face. 

 Moreover a mestome-sheath is plainly visible in this species, and is quite 

 thick-walled, in the larger bundles, while no such sheath is mentioned by 

 Professor Warming, as occurring in the material, which he examined. — 

 The structure of S. virginicus is thus very different from that of S. spicatus 

 (fide Volkens 1. c ), but is almost identical with that of S. pungens from 

 Europe. 



