The Natural History of British Grasses. 537 



to buy the cheapened produce, so long will this be a source of 

 weeds. Yet, so far as clean farming is concerned, we cannot 

 consider any as entitled to that name unless as well as destroying 

 weeds it also provides against sowing them. " Prevention " is 

 indeed better than cure, as weeding, however judiciously per- 

 formed, is sure to leave enough of prolific enemies to continue 

 the pest, so that it is the best, and safest, and, we think, the 

 cheapest cure. We have been tlms hard upon the lop, as it can 

 only be considered a weed, being an annual grass ; and, notwith- 

 standing the high position in which my friend Dr. Voelcker has. 

 placed it in Chemistry of Food, in respect to its feeding proper- 

 ties, which places it amongst grasses of superior quality, yet 

 cattle will not eat it if they can possibly avoid doing so, and hay 

 is always poor in which it occurs, which is not to be Avondered 

 at when the lop, for the most part, elects to grow in the most 

 impoverished soils. 



Bromus commutatus we can only view as a variety of B. mollis. 

 Its situation is that of low damp irrigated meadows, in which the 

 mollis is quite exceptional, though, when it does occur, it assumes 

 the drooping habit of the commutatus, and offers many interme- 

 diate states. Now, as we have watched the laying out of poor 

 pastures as irrigated meadows, we have observed that two or 

 three years is often capable of changing the B. mollis, which was 

 before alone, into commutatus. Of course it may be considered 

 that this was in virtue of that law of substitution of one species 

 for another Avhicli universally occurs on a change of soil, but we 

 incline to think that much of this where it occurs is due not 

 merely to this cause, but to real change of form, as the result of 

 an alteration of condition and circumstance ; and, as regards the 

 grass under consideration, our chain of evidence is nearly com- 

 plete in establishing this position, when it is stated that the 

 B. commutatus, from the irrigated meadows, most certainly in 

 cultivation in my experimental garden, has resulted in some fine 

 examples of B. seculinus, a form not before known there, and 

 therefore not liable to have led me into error, as miglit be the 

 case whore the diflerent varieties arc wikl natives. 



That B. arvensis, and perhaps other forms, may by cultivation 

 be shown to be varieties of an annual grass, of which B. mollis is 

 the coinmon type, is an idea which seems to be countenanced by 

 the protean forms of mollis and its congeners. 



B. crectns is a perennial brome, very partial to llmestcme 

 soils, and is one of the commonest grasses on the poor thin oolite 

 brashes, extending .ilong the whole of the Cotteswold chain of 

 hills, from I^ath to Chipping Campden ; it is no less prevalent 

 on the chalk range, and the quantity of lime which its ash con- 

 tains may have something to do with this preference. The per- 



