36 



JOURNAL OF THE 



analyses of the acid phosphate be calculated od a water -free basis, 

 so that they may be in the same condition, we can see at a glance 

 how the matter stands. 



TABLE L 



Shepard & Robertson have shown (in their very excellent pam- 

 phlet "on certain changes liable to occur in large heaps of acid 

 phosphate,") that if the mass maintains a high temperature (56° — 

 61°C.) for several weeks, there may result a considerable loss of 

 available phosphoric acid, (Available = Soluble + Reverted). But 

 this change probably does not begin for several days after mixing, 

 unless the crude phosphate used contained large quantities of iron 

 and aluminum. If a sample is drawn from the pan just before dis- 

 charging, put into a well closed bottle (corked), and cooled at once, 

 and allowed to stand for several weeks at the ordinary temperature 

 of the Laboratory, the reverse is the case. And the results obtained 

 by Shepard and Robertson from stuff ''which at no time after 

 drawing the first sample exhibited a higher temperature than 50°C," 

 confirm this statement. 



To test still farther the possibility of calculating out beforehand 

 the probable composition of an acid phosphate, another experiment 

 was conducted. The composition of the rock at this time was as 

 follows : 



Per cent. 



Moisture at 2I2°F 2.32 



Loss at red heat (COo restored) 3.03 



Insoluble silica 11.56 



Soluble silica 14 



Carbonic acid 3-33 



Phosphoric acid 26. 29 



Lime 38.55 



Oxide of iron (FeoOg) 3.35 



Oxide of aluminum 3.07 



