16 Mr. Woods on the Genera of European Grasses. 



Here, again, we may employ the compound terms Herbaceo-membranous 

 and Herbaceo-scariose to denote the intermediate states. The first occurs 

 in the glumes of Phalaris, and in the palese of Poa annua ; the second, in the 

 glumes of Anthoxanthum, and the glumes and paleae of Trisetum. 



Coriaceous. Thicker and firmer than membranous, as in the paleae of 

 Phalaris. 



Coriaceo-membranous would express an intermediate degree of consistency, 

 as in the paleae of Milium effusum. Herbaceo-coriaceous would denote a 

 difference of colour and texture rather than of substance ; we have it in the 

 paleae of Sclerochloa rigida and Triticum loliaceum. 



Horny. Harder than coriaceous. The paleae of Andropogon Allionii may 

 well deserve this title, at least when the plant is in seed, nor would it be mis- 

 applied to the paleae of Panicum, or the glume of Sorghum saccharatum. 



Fleshy. This term may be applied to the additional scales which indicate 

 the exterior florets in some species of Phalaris, and to those of the interior 

 floret in Melica, but perhaps not correctly to either glume or palea. 



The shape of the glumes and paleae, considered as flat surfaces, does not 

 appear to afford us any good generic characters ; but their being navicular, 

 compressed, concave, nerved, furnished with a marked midrib, or ribbed, are 

 particulars to be attended to. In this respect also I find considerable differ- 

 ence in the use of words. We have canaliculatce, carinatas, carinato-compressce, 

 naviculares, naviculari-carinatoe, concavce, concavo-compressce, naviculari-com- 

 pressae, involutce. Concave and carinate seem by most authors to be placed in 

 opposition ; but Smith unites in the same description concave and keeled. In 

 considering the substance of a plant, I understand the compound terms, such, 

 for instance, as herbaceo-membranaceous, to indicate an intermediate state ; 

 but in the foregoing examples it would be hard to say what idea was intended 

 to be expressed by a form between boat-shaped and heeled. Mertens and Koch 

 prefer considering the glume as convex instead of concave. 



The structure of the valve in this point of view consists of several particu- 

 lars unconnected with each other, which may or may not be combined, but 

 where in many cases we cannot with propriety speak of an intermediate form. 

 A navicular or boat-shaped glume is one so closed at the ends that it cannot 

 be opened into a flat surface. We find this in degree in the glumes of many 



