210 Gooch and Austin — Estimation of Manganese 



The most uncertain element in these experiments is the diffi- 

 culty, well- recognized at present, of getting the hydrous man- 

 ganous sulphate, upon which the experiments were made, in a 

 perfectly definite condition of hydration. 



Yolhard* subsequently studied the sulphate process, and 

 showed that manganous sulphate may be dehydrated, separated 

 from an excess of sulphuric acid, and brought into definite 

 condition for weighing as the anhydrous salt by careful and 

 protracted heating over a special device of his own — a ring 

 burner enclosed in a sheet-iron casing. Thus, on evaporating 

 and dehydrating a solution of pure neutral manganous sulphate, 

 Volhard obtained the results recorded in the following state- 

 ment : 



Residue of MnS0 4 left by evaporation and dehydration.. 0-1635 

 " after treatment with 3 drops of H 2 S0 4 and heating 



3 hours 0-1635 



" after heating 2 hours 0*1638 



" after treatment with 4 drops of H 2 S0 4 and heating 



2^ hours 0*1635 



" after heating 3 hours.. 0-1635 



Similar results were obtained on evaporating with sulphuric 

 acid and igniting in like manner an aqueous solution of man- 

 ganous chloride. Yolhard's recommendation of the method 

 has not secured for it the acceptance which its simplicity and 

 exactness would seem to demand — possibly because the periods 

 of ignition appear to be considerable and the manner of heat- 

 ing special. 



In our own experiments with the sulphate process we have 

 found that special apparatus is unnecessary, that the time of 

 treatment may be short, and that the process is in every respect 

 simple as well as very exact. We took for a starting point 

 manganous chloride prepared in the manner to be detailed. 

 An aqueous solution of the so-called pure manganous chloride 

 of commerce was boiled with pure manganous carbonate (to 

 throw out aluminum, iron and chromium), filtered, and pre- 

 cipitated with ammonium sulphide. The precipitate thus 

 obtained was dissolved in a very slight excess of hydrochloric 

 acid (to leave behind possible traces of nickel, cobalt, and cop- 

 per), the sokition was boiled to expel hydrogen sulphide and 

 precipitated with sodium carbonate. The manganous carbo- 

 nate thus thrown down was boiled repeatedly with successive 

 portions of. water, and washed until the washings were free 

 from chloride. The greater part of this purified carbonate was 

 dissolved in the least possible amount of pure hydrochloric 

 acid, the reserved portion of the carbonate was added, the mix- 



*Ann. d. Chem., cxcviii, 328. 



