S^* Liiiie ReqiiireiiK Ills of <'<rt<iiii Snils 



1)0 acid or neutral a considerable amount of the compound is directly 

 adsorbed and we have no means of determinins; this amount. The 

 experiments in the first part of this paper will suffice to demonstrate 

 this adsorption such neutral soils as the Rothamsted and Chelsea 

 samj^les exhibit marked powers for taking up calcium hydroxide. 

 Furthermore, where ammonia is used it is practically impossible In 

 obtain a clear filtrate or supernatant li(juid. As an alternative to 

 Tacke's method, Albert (."VJ) projxjsed one in which the absorptive 

 capacity of the soil for calcium, magnesium or barium was determined, 

 the latter being preferred on account of the lower dissociative power 

 of its salts. For the determination. 20 .'50 grms. of the air-dry soil 

 (according to acidity) are placed in a Jena glass flask together with 

 200 c.c. of distilled water; 50-100 c.c. of N/5 barium hydroxide solution 

 are run in from a burette and 10 grms. of sohd ammonium chloride 

 added. The flask is fitted immediately to a condenser and the contents 

 boiled for 20-25 minutes, the ammonia evolved being collected in N/10 

 acid and a titration made with sodium alizarin sulphonate as indicator. 

 Since the amount of the ammonia displaced from the ammonium 

 chloride is equivalent to the amount of barium hydroxide not absorbed 

 by the soil, this absorption or the soil acidity may be deteiniined. 

 The values obtained by the use of barium, calcium or magnesium 

 hydroxide were found to vary widely among themselves. According 

 to Siichting and Arnd (33), Albert's method is of little value since the 

 results are distinctly dependent on the period and intensity of boiling. 

 Lyon and Bizzell (34) have recently introduced a modification of 

 Albert's method. The reaction between soil and alkali does not occur 

 immediately and these workers recommend the addition of the barium 

 hydroxide solution about 60 minutes before that of the ammonium 

 chloride — the soil and alkali being kept at boiling point in the mean- 

 time. The procedure allows of two sources of error; the first is due 

 to the capacity of all soils to expel a quantity of ammonia from the 

 aniinonium chloride em])loy('d when the mixture is subjected tu In-at. 

 and Lyon and Bizzell suggest a correction for this — a blank deter- 

 mination being made with soil and ammonium salt but without alkali. 

 Ill some of our own work with this iiK'fhod the amount of iiiiiinoiiia 

 thus expelled by a soil deficient in lime and with the addition of the 

 prescribed quantity of animonium chloride amounted to 20 25 jier 

 cent, of the total produced on the addition of barium hydroxide as in 

 the standard method; the distillation of a further 50 c.c. increased 

 this error to over 30 per cent. \\\ the second place, all soils liberate 



