202 § 108. ANALYSIS OF SOILS AND HOCKS. 



in plants, let the whole stand several weeks, so that it 

 may settle together and assume a perfectly natural con- 

 dition. 



If the boxes, when put in place, are not surrounded by 

 grassy turf, the same plants as those cultivated in them 

 should be grown around them, so that there can be no 

 disturbing influences from fresh soil. 



When the seed is to be sown, put in more fresh soil if 

 it is necessary, in order to bring the surface up even with 

 the edge of the box. In sowing the seed, and in the sub- 

 sequent cultivation of the plant, the usual rules of good 

 culture should be observed as far as j^ossible, as respects 

 depth of sowing, distance apart of the plants, and other 

 matters. For experiments with the cereals, which are of 

 the greatest importance, oats and rye are best, since these 

 plants are less liable to disease, or to be destroyed by 

 birds. 



The following are a few of the great number of inter- 

 esting experiments that can be performed in these boxes 

 of prepared soil. 



Three or four soils may be prepared, differing great- 

 ly in the amount of sand and clay they contain, but 

 closely resembling each other as regards the lime and hu- 

 mus. Not only the quantity of the crop should be ob- 

 served, but also its quality^ such as, in the cereals, for 

 example, the proportion of grain, straw, and chaff, of 

 light and heavy grains, the specific gravity of the grain, 

 the number of the stalks and the degree of maturity at- 

 tained by them, the weight of the stubble and main roots, 

 and the proximate chemical composition of the different 

 parts of the plant. 



With such observations as these, combined with an ac- 

 curate knowledge of the composition of the soil, we 

 should soon learn whether any direct relation exists be- 

 tween the proportion of plant-food in the soil, that is 

 soluble in water or cold or hot liydrochloric acid, and 



