96 HORTUS GRAMINEUS WOBIUNENSIS. 



(which I doubt not. will happen, though at a remote period), 

 these comparisons between the value of the flowering- and 

 seed crops will possess their proper interest. The seed crops 

 of the natural grasses are in general, at least as far as my 

 observations have reached, left too lono- orowino- • the ereen, 

 or the withered state of the culms, is an uncertain criterion 

 to judge of the ri})eness of the seed of the perennial grasses, 

 though generally good for the ditTerent annual sorts. In the 

 greater part of the perennial grasses, the culms are far from 

 being either withered or dry when the seed is ripe, which is 

 determined, in almost all cases, by passing the spike or 

 panicle between the fingers ; if a portion of the seed separate 

 by this means, it will be found in the best state for collect- 

 ing ; it should, however, be suffered to remain in the ears 

 after being cut until the grass be perfectly dry. When the 

 grass is suffered to remain uncut till the culms are withered 

 and dry, and the seed separates on a slight touch, the grass 

 is rendered of little value, and the seed not improved in qua- 

 lity : the best part of the seed is either lost by devouring 

 birds or insects, or shaken out in the process of cutting and 

 collecting. I have made a number of experiments on the 

 seeds of grasses and other plants, by sowing at different de- 

 grees of what is termed ripeness, and the results went always 

 to prove the truth of the above remarks; also that diseased 

 or imperfectly-formed seed always vegetated best when sown 

 directly after being separated from the plant. An ounce of 

 this seed vegetated by this treatment in three different trials ; 

 the same seed, kept dry for two months only, did not vege- 

 tate by the like means employed in sowing, or any other 

 that I attempted. 1 have repeated experiments of this 

 nature with much interest and care, but the principle still 

 remained unaltered. 



The produce of latter-math is 6,125 lbs. per acre. 



From the above results it appears, that this grass is much 

 inferior to i\\Qfestuca duriuscula in the quantity of its pro- 

 duce, as well as in nutrient qualities ; which will be evident 

 on a comparison of these properties, as mentioned in the fore- 

 going details. It springs rather earlier than the hard fescue, 

 and also rises better after being cropped, but not, apparently, 



