190 AGRICULTURE. BOOK I. 



and consequently ought to be removed, if not previously, at 

 least immediately after the period when they come into flower ; 

 that some vegetables send out their roots horizontally along the 

 surface, and are easily killed by being planted too deep in the 

 earth ; as the mentha, tanacetum, and couch-grass, or triticum 

 repens ; that some sorts are destroyed by rendering land firm 

 and close, as the common nettle ; and others, by rendering it 

 soft, as the common ragweed, or hieracium, Sec. ; that good plants 

 are often produced by making land firm, as the poa trivialis ; 

 others by simply laying it dry, as the white clover, &c. &c. 



With respect to discovering the nature of soils by the proper- 

 ties of plants, it can be done with greater certainty by this than 

 by any other branch of science. Growing crops may fre- 

 quently deceive; but when the natural plants or weeds are 

 discovered, they are almost infallible indications of the nature 

 of the soil. Mr. Kent, in his hints on landed property, has 

 very judiciously mentioned several plants which indicate cer- 

 tain qualities of soil ; to which any person practically ac- 

 quainted with agriculture, and accustomed to make botanical 

 excursions in the fields, might add many others ; for example, 

 in the genus plantago, the species lanceolata indicates a deep, 

 free, or sandy loam ; the p. major a firm good loam ; the p. co- 

 ronopus, linearis, maritima, arenarious, gravelly, and poor 

 soils. The tussilago petasites near the margin of rivers is a sure 



