516 



Transactions. 



pounds per acre, the range is from 1,900 lb. to 2,400 lb.— that is, the highest 

 figure is more than 25 per cent, greater than the lowest. It was recognized, 

 however, that the determinations which failed to agree had been made under 

 different conditions as to volume and strength of bicarbonate solution, 

 and so it seemed necessary to make sets of determinations simultaneously 

 on one soil under uniform conditions. Results of such are shown in 

 Table D. 



Table D. 



In sets B and C 400 c.c. of solution was used ; in set A only 300 c.c, 

 the strength of which was also rather less than in the case of the other 

 sets. 



As a further test, three samples of soil from field 16 were treated simul- 

 taneously- — A with 300 c.c. of solution of strength 0-02 N ; B with 400 c.c. 

 of the same strength ; C with 400 c.c. of strength 0-014 N. The lime re- 

 quirements indicated in these three cases were — A, 0-11 per cent. ; B, 0-12 

 per cent. ; C, 0-05 per cent. It seems, therefore, that the volume of solu- 

 tion may be varied within fairly wide limits without affecting the result, 

 provided that the solution does not at any time fall below a certain 

 concentration. This is an obviously suggestive point, which merits 

 investigation. 



IV. 



In one experiment a positive, though small, lime requirement was indi- 

 cated for a limestone-derived soil, and this suggested the possibility of a 

 weakening of the solution (which corresponds to a lime requirement) from 

 physical as well as from chemical causes. To test this idea a series of trials 

 was made. Four soil-samples that had already been in contact with solution 

 for twenty-four hours in connection with previous trials, and which were 

 now presumably satisfied as regards their lime reqiurements, were filtered 

 from their old solutions, and treated again with fresh solution. A sample 

 of sand was prepared by treating alternately with concentrated HCl and 

 strong ammonia solution, washing thoroughly, and separating a uniform 

 sample by sedimentation. About 9 grams of this sand was treated in the 

 same way as the soils. Another bottle contained a soil derived from lime- 

 stone from Waikari ; and, lastly, a bottle of the bicarbonate solution 

 without any soil at all was put through the same processes as the othei's 

 samples of this series. The results are given in Table E, the first four 

 soils being which, having been previously treated, were presumably already 

 saturated. 



