Holm, Some new anatomical eharacters for certain Gramineae. 12^ 



midrib, and also accompanies the lateral nerves on the leptome- 

 and hadrome - side of these. A dense palisade - tissue sur- 

 rouuds the mestome-bundles and also oceupies the Spaces between 

 these, underneath the bulliform eells, excepting near the midrib, 

 where this spaee is taken up by a few layers of colorless tissue. 

 The cells of the palisade-tissue are arranged radialiy towards the 

 Center ot the mestome-bundles and border directly on a thin- 

 walled, large-celled, green parenchyma-sheath, inside of which is 

 a mestome-sheath of thick-walled cells. There is also in the larger 

 bundles a layer of thick-walled mestomeparenchyma between the 

 leptome and hadrome. The outline of the bundles in trarsverse 

 section is oval in the larger ones, and orbicular in the smaller, 

 the latter predominating 



On the whole, the general leaf-structure of these associates 

 of Aristida displays the same and constant development of a single, 

 mostly large-celled parenchyma-sheath , besides a mestome-sheath 

 inside this. — Furthermore the mesophyll consists of palisades 

 arranged radialiy around the mestome-bundles, and this cell-form 

 is the prevalent in these species ; though Glyceria is exceptional 

 in having not only palisades, which border directly on the paren- 

 chyma sheath, but also some layers of much shorter and nearly 

 roundish cells with Chlorophyll between the ribs, underneath the 

 bulliform cells. A colorless tissue was observed in most of the 

 species; Glyceria and several species of Sporolohus (S. airoides^ 

 argutus, hrevicalyx, cryjjtandrus, indicus etc.) lack this besides that 

 it is only developed as a small group, one on each side of the 

 midrib in Scleropogon, or only above the midrib, as was observed 

 in Sporobolus tenacissimus. But otherwise the structure seems very 

 uniform and quite similar to that of the allied genera, besides 

 also exhibiting several points in common with Aristida. 



Deductions. 



Contrary toour expectation we have not succeeded in de- 

 tecting the double parenchyma-sheath in any of the genera, that 

 are allied or associated with Aristida, and it constitutes, no doubt, 

 a generic character. Its absence from some of the Aristidae does 

 not seem to indicate that such modifications in structure are induced 

 by diversity in environment, but it seems more probable that 

 these species, in which there is only one sheath developed, should 

 be removed from Aristida proper, and especially because their 

 morphological eharacters are so very distinct. When we at first 

 noticed the two Chlorophyll - bearing sheaths in Aristida fasci- 

 culata, we thought that the innermost might correspond to the 

 one whieh Haberlandt (1. c.) detected inside the mestome-sheath 

 of Cyjjerus, but Professor Schwendener has kindly informed. 

 "US that he is more inclined to consider them as histologically 

 distinct, even if their function be identical. The outermost of 

 these sheaths compares, on the other hand, with the ordinary 

 parenchyma-sheath, known so well from the niestome-bundles in 



