ELISHA MITCHELL SCIENTIFIC SOCIETY. 117 



from the dry to the moist condition of the soil, all or nearly all 

 have to be partially neglected and only in a crude way can thev 

 be imitated. 



And again, the proportion of the soluble phosphoric acid of 

 the superphosphate to the quantity of the soil in the experiment 

 cannot be made to agree with that in nature; this also, in a gen- 

 eral way, must be approximated. 



The result, when reached, will express at least not absolute 

 but relative results. 



The plan pursued was to mix an acid phosphate with soils of 

 different composition, and to observe by analysis the changes 

 resulting from such a mixture after the lapse of time. The main 

 difficulty experienced in the outset was that of procuring iden- 

 tical portions for analyses. One large heap mixed in the proper 

 proportions and sampled at stated times would not answer the 

 conditions; for the sampling itself might not contain the proper 

 quantity of the soil and fertilizer, and so prove a fruitful source 

 of error. 



The difficulty was met by mixing at the same time many little 

 heaps in the same proportion and in such quantity that the whole 

 heap could be analyzed at stated intervals. 



The details of the experiments were as follows: 



A. The acid phosphate used was freshly prepared from S. C. 

 Rock, carefully ground to pass a 60 mesh seive, and carefully 

 mixed. Its analysis (using the method of Official Agricultural 

 Chemist for 1886-'87, where the insoluble phosphate is treated 

 for 30 minutes at a temperature of 65° C, shaking at intervals 

 of 5 minutes) resulted as follows (Table IV): 



TABLE IV. 



ANALYSIS OF ACID PHOSPHATE. 



Moisture. 



Total Phos. Acid. 



Sol. Phos. Acid. 



Insol. Phos. Acid. 



, . I '.v fusion 15.26 Assoc. Method 12.63 Assoc. Method 0.92 



By Assoc. Method... L5.29 Vssoe. Method 12.82 Assoc. Method 0.87 



I Average 15.27 Average 12.72 Average I 



B. The soils chosen were four in number and of varied char- 

 acter. The original field sample was air dried, crushed with the 



