25i AGRICULTURAL MUSEUM 



The other pliysical properties of soils should likewise 

 be examined before the aiialysis is made, as they de- 

 note, to a eertain extent, their composition, and serve 

 as o^nides in direetiiiii, tlie experiments. Tlais siliceous 

 soils are generally rough to the touch, and sciateh glass 

 when rubbed upon it; aluminous soils adhere strongly 

 to the tongue, and emit a strong earthy smell when 

 breathed upon ; and calcareous soils are soft, and much 

 less adhesive than aluminous soils. 



V, Mode of ascertaining the Quantity of Water 

 OF Absorption in Soils. 



Soils, though as dry as they can be made by continued 

 exposure to air, in all cases still contain a considerable 

 quaiitity of water, which adheres with great obstinacy 

 to the earths and animal and vegetable matter, and can 

 only be driven oft" from them by a considerable degree 

 of heat. Tlie first process of analysis is, to free the 

 given weight of soil from as much of this water as 

 possible, without in other respects aSecting its composi- 

 tion ; and this may be done by heating it ten or twelve 

 minutes ovei- an Argand's lamp, in a bason of porcelain, 

 to a temperature equal to 300 t Fahrenheit ; and in case 

 a therm(»meter is not used, the proper degree may be 

 easily ascertained, by keeping a piece of wood in con- 

 tact with the bottom of the dish ; as long as the colour 

 of the wood remains uisaltered, the heat Js not too highj 

 but when the wood begins to be charred, the process 

 must be stopped. A small quantity of water will per- 

 haps remain m the soil even after this operation, but it 

 always affords useful comparative results ; and if a high- 

 er ieini)(;rature were emi>loyed, the \egetable or ani- 

 mal mailer would undergo decomposition, and in con. 

 sequence the expe, iment be wholly unsatisfactory. 



The ioss o! weight in the ju-ucess siiouli be cartiully 

 noted, and ivhen 111 fuar hundred g.aias oi soil it i cach- 

 es as hign as 50, ttu so. I may be considered as in Uie 



SI . V-- ai •x^'fiaiTts ia which this proce.s b« l.«-. carrlc? on by dist.ll.-.tion. Il.ave foou* 

 tht waiur mat cam. 07« jure, and no icasMc quantity vi volatile maaer ^^^t j^roducerf. 



