ELISHA MITCHELL SCIENTIFIC SOCIETY. 29 



exhibited by a cloudiness that was greates^t iu (>, nearly as much in 

 3, next in 2. next in 4 and very faint in 5. They were all slight, but 

 tend to show that the addition of oxalate m ly bp nf'ces.sary for an- 

 other reason than, in the language of the eoiiiuiittee, ' to facilitate 

 the conversion of any nitrates of potash which may be present into 

 carbonates upon subsequent evaporation and ignition."] After add- 

 ing the ammonium oxalate the solutions were evaporated to dryness 

 in platinum dishes, volatile matters (amuionia salts) expelled by 

 heating below red heat and the residues taken up in small amounts 

 of hot water, and filtered through small filters in porcelain dishes; 2 

 drops HCl were added and then 10 c. c. PtCl^ solution; evaporated 

 until mass solidified in cooling; took up in 85 per cent, alcohol and 

 finished as laid down in a. 



This method of weighing I regarded as superior to all others in 

 point of accuracy, and hence, contrary to the committee's recom- 

 mendations, I preferred to vary from the method in this slight par- 

 ticular, inasmuch as I do believe it advisable to wash the precipitate 

 out with water and weigh the filter again afterwards. Furthermore 

 I believe that the nicety of adding the reagents, drop by drop with 

 constant stirring is in commercial analysis unnecessary. 



Results. — In the chemically pure K^SO^ determinations: 



a. gave K^PiCIh, 0.2S01 gram, ^ KoO, 54. oS per cent. 



^. " " 0.2805 gram, = " 54-i6 



e. " " 0.2829 S'"^'''''> = " 54-62 " 



The per cents are seen to be quite close together, but it is to the 

 weights of K^PtCly that we should look, since the per cents vary 

 according to the amount of substance taken — in this case very small, 

 0.1 gram. The per cent, in c is much higher than the theo- 

 retical, but the smallness of the amount worked with must be borne 

 in mind. The weights of K^>PtCI,; are very satisfactorily near to- 

 gether. A tendency to high result (c) is shown probably from ex- 

 tended manipulation introducing foreign substances into the solu- 

 tion. 



1.0 gram of substance is the amount taken in potash determina- 

 tions in commercial fertilizers and was the amount taken in the 

 committee's samples in this work, 3, 4, 5 and 6. Now suppose the 

 above three weights of K^PtCl,;, instead of coming from the very 

 small 0.1 gram, came from 1.0 instead. The corresponding per 

 cents of K2O would be 5.41, 5.42 and 5.46, which are very satisfac- 



