12 PROCEEDINGS OF THE AMERICAN ACADEMY 



trichloride on turmerol, but much less neatly. Phosphoric penta- 

 chloride, on the other hand, seemed to add chlorine also, as the product 

 from its action on turmerol contained about 15 per cent of chlorine. 

 "We have postponed the study of this substance for the present. 



Tlie formation of the turmerylchloride described above indicates 

 that turmerol is an alcohol, and to confirm this view we next treated 

 the turmerylchloride with various reagents, and found that its chlorine 

 was removed, in part, at least, by boiling water or by alcoholic solu- 

 tions of sodic acetate, potassic cyanide, or ammonia, — a substance being 

 formed in each case with the characteristic smell of the class to which 

 it should belong. The regenerated alcohol and the acetate we were 

 unable to purify, while the nitrile and its corresponding acid and the 

 amine were formed in very small quantities and did not give crystal- 

 line compounds, so that their further study, which we had at first 

 hoped would establish the formula of turmerol, did not jiromise to 

 repay the labor. 



To still further confirm the alcoholic nature of turmerol, we prepared 

 the sodium compound as follows, — 



Action of Sodium on Turmerol. 



Some of the turmerol mixed with high boiling ligroine was warmed 

 with sodium in a flask with a return cooler for twelve hours, during 

 which time hydrogen was given off ; the liquid was then filtered, and 

 the ligroine evaporated off from the viscous filtrate. 



I. 0.1456 g. of substance gave 0.0374 g. of Na2S0^. 

 II. 0.9820 g. gave 0.2342 g. of Na^SO^. 



Calculated for C,QlI,,0Na. Found. 



I. n. 



Sodium 7.82 8.32 7.72 



The substance forms a viscous semi-solid mass. 



Sodic hydrate seemed not to act upon turrperol. 



An attempt to convert turmerol into its acetate by heating it 

 with acetic anhydride and sodic acetate on the water-bath gave 

 a product having the same smell as that from the reaction of sodic 

 acetate on turmerylchloride, but no means could be found of separ- 

 ating it from the unaltered turmerol, as it was decomposed by distil- 

 lation even in vacuo. We accordingly turned our attention to the 

 ethers of turmerol, and selected for study the isobutylether, as the 

 difference between the percentage composition of this and that of 

 the alcohol would be greater than in the case of the methyl- or ethyl- 

 ethers. 



