179 



stances, that it is difficult, for one accustomed to the close investigation 

 of specific characters, to believe that the one before us has claim to be 

 considered otherwise than as a very luxuriant form of its congener. 

 Mr. Bentham assures us that " in North America, where it is frequent, 

 it passes gradually into the more common form ;" and I have seen 

 Southampton specimens which it would be impossible to regard in any 

 other light than as intermediate varieties. Dr. Bromfield mentions S. 

 altemiflora as having been met with on the banks of the Adour near 

 Bayonne, on the coast of France. 



The rank odour above referred to is by no means peculiar to this 

 American species or variety, whichever it may be, being not at all 

 uncommon to plants vegetating in similar localities, and, though un- 

 noticed by authors as a property of 8. stricta, that grass will be 

 found, when first collected, to be almost equally unpleasing to the 

 nostrils. 



Genus 48. CYNODON. Dog's-tooth Grass. 



Gbn. Chah. Inflorescence a digitate, umbellate, or racemose cluster 

 of spreading, linear spikes. Spikelets nearly sessile, equidistant 

 in a single row on one side of the rachis, laterally compressed, 

 one-flowered (occasionally with a superior rudiment of a second 

 flower). Glumes two, nearly equal. Pale» two, compressed ; 

 the lower boat- shaped, eventually hardening around, and enclosing 

 the caryopsis (fruit). Styles distinct. 



A small, but widely distributed genus, separated from Panicum, to 

 which the typical species, Cynodon Dactylon, was referred by Linnaeus. 

 Named from the Greek cyon, a dog, and odous, a tooth, adopted by 

 Persoon, probably as characteristic of the arrangement of the leaves in 

 the creeping, barren shoots, as seen spreading over the surface of the 

 soil. The rudiment of the second flower in the spikelet is usually a 

 mere filament, proceeding from the outer base of the external palea, 

 and occasionally either thickened or bearing a minute scale at its 

 apex. From its position this filament may be regarded as an extension 

 of the floral axis, and the terminal scale as the rudimentary palea of 

 another flower, which is therefore superior or above the perfect one. 

 This explanation is important in a physiological point of view, but 

 unnecessary to the recognition of the genus. 



Cynodon dactylon. Creeping Dog's-tooth Grass. Plate CXLII. 



Clusters of from three to six spikes. Outer palea longer than the 

 glumes, smooth, slightly ciliated on the keel and margins. Leaves 

 downy beneath ; those of the creeping barren shoots flat, spreading ; 

 those of the culms usually folded inward. 



