ELISHA MITCHELL SCIENTIFIC SOCIETY. 



29 



I will add that the loss of water in A. and B. is rather excessive, 

 part of which being due to artificial drying as the sample was too 

 wet to pass the fine seive. 



It will be seen that the decrease of Insol. PoO-, in D. is very slight, 

 only .12 per cent, from 20 to 100 mesh. I was at a loss to understand 

 this remarkable deviation from the other results until after the de- 

 terminations were completed, when I found from the manufacturers 

 that the basis of the phosphate was a natural guano instead of a 

 manipulated superphosphate. The reason then was plain — ammo- 

 nium citrate solution acts no quicker on fine than on coarser parti- 

 cles of raw phosphate. 



For the sake of comparison other results are calculated and given 

 in Table II. 



TABLE II. 



From the last table we see that incident to the loss of water and 

 decrease of Insol. PgO- is an increase of Total and consequently an 

 increase of reverted and available P^Og, amounting in some cases 

 to nearly one per cent. 



To summarize then we have for the results of pulverization of fer- 

 tilizer samples : 



1. The more perfect mixing of the separate ingredients, and the 

 consequent susceptibility to better and fairer sampling. 



