82 PROCEEDINGS OF THE AMERICAN ACADEMY 



supersaturation, and arsenous oxide separates from the solution on cool- 

 ing. The genei-al statements made concerning the arsenoso-tungstates 

 apply to these salts also. The general formula is 



m MoOo . 11 AsoOg . p RaOj 



in which n is usually greater than unity. The salts are often beauti- 

 fully crystalline. 



Analytical Methods. — Arsenous oxide may be determined in these 

 salts by oxidizing the solutions in presence of an excess of an alkaline 

 carbonate by means of bromine, and precipitating as ammonio-magnesic 

 arsenate. The pi'ecipitation must be repeated a second time, but the 

 final treatment with ammouic sidphide, which I have recommended for 

 the separation of phosphoric and molybdic oxides, is of course inadmis- 

 sible. By far the best method of determining the arsenic in these salts 

 is titrition with iodine after addition of an excess of an alkaliue bicar- 

 bonate. It is not possible to determine the oxides of molybdenum and 

 arsenic together by precipitation with mercurous nitrate, since no 

 amount of care suffices to prevent the volatilization of arsenic in the 

 subsequent ignition. For the same reason, it is also very difficult to 

 determine the water of crystallization by the ignition loss, as fumes of 

 arsenous oxide are given oflF at temperatures which are much below 

 that at which sodic tungstate fuses. The degree of facility with which 

 this takes place varies with the salt to be analyzed. As regards the 

 determination of the water, I have found it best to heat gently in a 

 jiorcelain crucible, suj^ported over a small iron cup heated from below, 

 but even in this case the results are never very satisfactory. Arsenous 

 oxide cannot be separated from molybdic oxide by hydric sulphide, 

 since molybdic sulphide is always precipitated at the same time, even 

 when phosphoric acid has been added in excess. I have, however, 

 found that fairly good determinations of molybdic oxide may be made 

 in many cases by dissolving, boiling with a small excess of am- 

 monia, adding a little ammonic chloride, and finally precipitating 

 with baric chloride. The baric molybdate may then be ignited and 

 weighed upon an asbestos filter. The results, if not very accurate, 

 are at least sufficiently close to be of great service as verifications of 

 formulas. 



8 : 2 Baric ArRennso-molyhdafe. — A boiling strong solution of 14 : 6 

 ammonic molybdate dissolves arsenous oxide in large quantity, and 

 very readily. The solution has usually a yellowish-green color from 

 slight reduction of the molybdic teroxide. After filtration and evapo- 

 ration uj)on a water-bath, it deposits an abundance of octahedral crys- 



