92 JOURNAIy OF THE 



cooled somewhat, and an equivalent amount of sodium 

 ethylate added, and rapidly cooled to the same tempera- 

 ture. The precipitate was filtered off by means of pump 

 and washed with absolute alcohol, avoiding- all exposure 

 to the air possible. It was then transferred to the pre- 

 cipitating- flask and an equivalent amount of eth}^ iodide 

 added. 



The mixture in the flask provided with a reflux con- 

 denser, was g-radualh^ warmed up to boiling*. The 

 chang-es first noted were the formation of a dark brown 

 deposit on the bottom of the flask at about 70° C, a 

 coloring- of the liquid, and at the same time a diminu- 

 tion of the precipitate. Finally, at the boiling point 

 of the mixture, the precipitate appeared to become 

 sticky, and to collect into one mass, instead of being; 

 flocculent, and to gradually get smaller and smaller, 

 both going into solution and coloring- it a dark brown 

 and melting- down to a semi-syrup on the bottom of the 

 flask. 



The, time, in this test, for the chang-e was one hour; 

 much longer than in the former experiments. The 

 liquid in the flask was divided into two portions, one of 

 which was boiled in a flask with a reflux condenser for 

 several hours, and no chang-e was observed. 



It reduced Fehling-'s solution and had a strong- smell 

 of eth^d iodide. The portions were now combined and 

 submitted to fractional distillation. 



About half of the amount of eth}^ iodide used was 

 recovered in the fraction coming- over between 74° and 

 780 C. 



The alcoholic fraction emitted still a strong smell of the 

 iodide. Hence it seemed that the ethyl iodide had 

 played no part in the change undergone by the sodium 

 glucosate. The residue left in the flask from the experi- 



