HORTUS GRAMINEUS WOB U KN ENSIS. 131 



pratensis (meadow oat), Festuca duriuscula (hard fescue), Poa tri- 

 vialis (smooth-stalked meadow-grass), Poafertilis (fertile meadow- 

 grass), Poa nervata (nerved meadow-grass), Trifolium medium 

 (cow clover), Trifolium repens (Dutch, or white clover), Agrostis 

 stolonifera (stoloniferous bent, or fiorin), and Agrostis palustris 

 (marsh bent). 



The seeds of the six first-mentioned grasses being much larger 

 than the others, were first mixed and sown, and covered with the 

 rake ; the rest of the seeds were mixed and sown, without any 

 other means of covering but that afforded by the roller, which was 

 liberally employed till the surface was perfectly level and consoli- 

 dated. This was effected on the 23d of August, 1813. 



The seeds of all these grasses vegetated before the first week of 

 October, except the seed of the Vicia sepium, w r hich did not vege- 

 tate till the autumn of the succeeding year. Before the frost set 

 in, these seedling grasses had a top-dressing with compost of 

 rotten dung, lime, and vegetable mould, laid on in a fine and dry 

 state, after which the ground was again well rolled : in the month 

 of February this operation was again repeated, when the ground 

 was sufficiently dry to admit of it. The plants sprang earlier than 

 those of the old pasture (a circumstance common to young plants 

 in general). In April, the weeds which had accompanied the top- 

 dressing were carefully cleared away, and the rolling was repeated, 

 to keep the surface compact. The plants grew vigorously till a 

 continuance of unfavourable weather in the end of June checked 

 their growth. In the first week of July the produce was cut and 

 weighed ; it amounted to one-eighth more than the produce of 

 the ground in its original state, but which had been fed off with 

 sheep in the spring ; the after-math of the seedling grasses, how- 

 ever, weighed one-fifth less than that of the natural pasture. 

 A very slight top-dressing was applied in the month of November, 

 and the whole was then well rolled ; this operation was continued 

 at favourable opportunities till April last (1815). The grass was 

 cut and weighed in the first weeks of June and August, and again 

 in the middle of September; the total weight of these three crops 

 exceeded that of the old turf exactly in the proportion of nine to 

 eight. 



It is therefore evident, that the results of the two modes of ex- 

 periment here adopted perfectly agree in confirming the opinion, 

 that a five years' course of the more impoverishing annual crops 

 may be taken from land of the nature before described, without 

 unfitting it for the reproduction of the superior natural grasses. 



