1864'.] Prof. A. H. Church on certain Ethylphosphates. 523 



These numbers correspond to the formula (C^ Hg)3 Fe^ SPO^ + S Aq, as 

 may be seen in the comparison given below of the theoretical and experi- 

 mental percentages. 







Theory. 





Experiment. 











I. 



II. III. 



IV. 



Ce ... 



. 72 



13-38 



13-19 







H,, ... 



. 21 



3-91 



3-92 







Fe, ... 



. 112 



20-82 





20-82 20-56 





Pa ... 



. 93 



17-28 







17-28 



0, ... 



, 240 



44-61 











538 



100-00 









In analysis V, 10-17 per cent, of water were lost by drying the air-dried 

 salt at 150°: the formula above given requires 10-03 per cent. If we 

 allow the formula fe CI to express the molecule of ferric chloride, giving to 



28 X 2 



the iron in it the atomic weight — - — = 18-67, then the ferric ethylphos- 



o 



phate may be written 



C, H. ] 

 ■ fe' lPO, + Aq, 

 fe' J 



Few chemists would now admit such an expression to be anything more 

 than what may be termed an equivalent formula, comparable with that of 

 ethylphosphoric acid itself, yet representing one-third only of the true 

 atomic weight of the iron compound. It may, however, be worth while to 

 consider whether there be any mode of arriving at a decision concerning 

 the formula of the ferric ethylphosphate — whether the above simple ex- 

 pression be admissible, or the more complex form 



Fe^-^ I 



Fe^-^ ^SPO^+SAq. 

 Fe'-' I 

 Fe'-^ J 



It appeared to me that, if the latter expression be the true one, we ought 

 to be able to replace ^th or |-ths of the iron present by another metal : if 

 the simpler formula be correct, then any other replacement but that of 

 j would be impossible, unless indeed we suppose that the very chemical 

 process made use of to effect the replacement causes a coalescence of three 

 atoms of the original salt, in order that one more complex atom of the new 

 mixed compound may be thereby constructed. 



With this object in view, several experiments were devised. A solution 

 saturated at 60° and containing a known proportion of ferric and aluminic 

 ethylphosphates was brought to the boiling-point, and the salt thus 

 separated removed by filtration. In other cases absolute alcohol was 

 added to the warm concentrated solution till a part only, often but a small 

 part, of the salt was precipitated. By these methods, and by the action of 



