150 PROCEEDINGS OP THE AMERICAN ACADEMY. 



Determination of the 3folecular Weight of the Oxide. 



This determiuatiou was made to find out whether the oxide was 

 derived from one or more molecules of the benzoylacetal. The boilinw 

 point method of Beckmann was used with benzol as the solvent, and the 

 following results were obtained : — 



These results show that the oxygen atom does not bind together two 

 benzol rin^s. 



'&" 



■ Action of Sodic Hydrate on the Oxide C,C1oOCH30(OCOCbH5)oOCH3. 



The oxide did not dissolve in a cold solution of sodic hydrate, but 

 when the two substances were allowed to stand together for twenty 

 hours, it was found that a small quantity of sodic chlorauilate had sepa- 

 rated out. 



If, on the other hand, the oxide was boiled with sodic hydrate, it was 

 decomposed rapidly even when a dilute solution was used. 1.5 grams 

 of the oxide were boiled for half an hour with a solution of 5 grams 

 of sodic hydrate in 60 c. c. of water ; at the end of this time the oxide 

 had disappeared completely, and the liquid had taken on a dark purple 

 color. It was filtered while still hot, and the filtrate on cooling depos- 

 ited crystals, which looked like sodic chlorauilate. To determine their 

 composition they were collected on a filter, and, after washing with 

 alcohol and a little water, recrystallized from hot water. The crystals 

 were then dried over sulphuric acid, and finally at 110°-120°, when an 

 analysis gave the following result : — 



0.1707 gram of the salt gave 0.0949 gram of sodic sulphate. 



Calculated for C„Cl2(0Na)202. Found. 



Sodium 18.18 18.02 



The product therefore was sodic chlorauilate. 



