Atkins — Errors in Hydrogen Ion Determinations of Soils. 345 



Table V. 



pR VALUES OF MLTEU-PAPER, SPOT TESTS. 



Paiier. pH. 



Whatman, No. 1, 11 em., old batch, ... ... ... 7-6 



Do. do. new batch, ... ... ... 6- 8-7-0 



Do. No. 1, 12-5 cm., old batch, ... ... ... 7-4-7-6 



Do. No. 4, „ „ ... ... ... 6-8-7-0 



Do. No. 40, 11 cm., HCl and HF, ... ... 4-4-4-S 



Munktell 's No. 0, 7 cm., HCl, ... ... ... 4'4-4-8 



Do. „ 9 em., HCl, ... ... ... 4-4^-8 



Schleicher and ShuU, No. 589, 7 cm., HCl and HP, ... 4-4-4-8 



Chardin, for agar agar, ... ... ... ... 4. 4-4.3 



J. Green, extraction thimble, ... ... ... ... 5-2 



It was furt.hennore found that washing was not effective in lessening the 

 action of acid-extracted paper in absorbing acid. Thns when 100-150 c.c. of 

 distilled water was filtered through Munktell's No. paper it came out at 

 pH 5-65, namely a little acid had been taken up by the paper. Peat extract, 

 found to be at -pH 4-55 when centrifuged, was then filtered through both 

 washed and unwashed paper, and the filtrates (10 c.c.) were both at pH 4-8. 

 With unextracted paper there was a noticeable improvement through washing, 

 but the error still remained very great, viz. pH 1-45. This i)aper (Whatman 

 No. 1) when washed as before v/ith 100-150 c.c. of water at pH 5-6 reduced 

 the acidity of the latter to pH 62, having evidently parted with alkali. When 

 peat extract at pH 4-55 was filtered through washed and umvashed paper the 

 10 c.c. filtrate was at pH 6-2 for the former and pH 60 for the latter. 



Indicators. 



Those of the Clark and Lub's series have been found by the writer to bo 

 very satisfactoiy and stable in solution, with the exception of methyl red. 

 Cohen (1922) has introduced brom cresol green as a substitute for methyl 

 red, and its use is to be preferred. It is like the other sulphone phthaleins 

 in being unabsorbed by peat and plant tissues, whereas methyl red is rapidly 

 taken out of the solution, quite apart from being decolorised by bacterial action. 



The indicators should be used in their half transformed condition, phenol 

 red for example being neither a. deep red nor j^ellow, Ijut a faint j^ink. This 

 is of importance when dealing with lightly buffered soil extracts and natural 

 waters. The use of alcohol as a solvent should be avoided where possible, 

 though for methyl red it is certainly convenient. Alcohol is apt to undergo 

 oxidation, producing traces of acetic acid, and so it is usually faintly acid; 

 acetic acid and sodium acetate act as buffers, so their addition in the indicator 

 solution should be avoided. For this and other reasons when there is a difference 

 between the pH vakies given by a soil extract with methyl red and brom 

 cresol purple the value shown by the latter should be accepted or the newly 

 introduced brom cresol green used. 



In the course of a discussion on certain of the points raised in this paper. 

 Dr. A. F. Joseph suggested that the alkaline soils of the Sudan, in which the 

 clay was deflocculated by the alkali, might not be satisfactorily cleared by 

 the centrifuge. He accordingly sent a sample for trial, which contained 

 296 per cent, of clay, and was extremely alkaline, pll 100, as determined 

 electrometrically. 



