42 JOURNAL OF THE 



c. c. of filtrate; neutralize with NH4HO; add about 15 grammes 

 dry ammonium nitrate, and to the hot solution add molybdic solu- 

 tion sufficient to precipitate all the PoO., present. Digest at about 

 65" C for one hour, filter and wash with ammonium nitrate solution. 

 Now in the first place the drying down vvith magnesium nitrate solu- 

 tion must of necessity be done over the water bath ; it must, moreover, 

 be transferred to the air bath and the drying finished at a higher 

 temperature, else there will be loss from spitting when it comes to 

 be ignited, however gently. It next requires a long ignition to de- 

 stroy organic matter, and if, as the method suggests, moistening 

 with HNO, becomes necessary the drying process must be repeated. 

 It is quite difficult, also, to decide on a convenient form of capsule 

 to be used. I tried two kinds in two different sets of experiments, 

 one kind being simply an extra large best porcelain crucible, and 

 the other being a capsule more on the order of a fiat-bottomed 

 evaporating dish. I made eight experiments with the crucibles and 

 four with the " capsules." In every single instance in the case of 

 both kinds they cracked at the "gentle ignition" stage. However 

 carefully and slowly I applied the heat in each case, the ominous 

 " crack " was heard, and this too some little while after they had 

 attained their maximum temperature. Such wholesale breakage 

 makes the method very expensive, and I cannot see how it can be 

 avoided. In our fusions we always used platinum crucibles with 

 economy. Now the ignited mass must be treated with fuming HCl. 

 Whatever be the form of capsule used the violent chemical action 

 caused by pouring HCl on this mass of MgO causes serious loss by 

 spitting and it is not convenient to cover the capsule when the HCl 

 is added. Afterwards, also, when digested at a heat however gentle, 

 serious loss may again occur from spitting (to be avoided with much 

 inconvenience) when the first bubbles of HCl gas are given off. And 

 then in case one has a broken capsule, as I did in every instance, 

 the capsule and contents must be enclosed in some other containing 

 vessel (as a beaker) and there treated with HCl under cover of a 

 watch glass. This necessitates the addition of a very large quan- 

 tity of HCl in order to cover the crucible and contents, and the 

 fumes from this HCl, in the case of a dozen or more analyses going 

 on at once, render the atmosphere of the laboratory very disagree- 

 able despite the use of a hood. 



Such were the numerous practical difficulties I met with in the Phil- 

 adelphia method. 



For my experiments I selected four samples of commercial fertil- 



