OF ARTS AND SCIENCES. 285 



Here again the analysis gave the desired result, but our auilid melted 

 at 143°, whereas according to Fischli * the anilid of paratoluylic acid 

 melts at 139°, according to Bruckner t at 140-141° ; we felt, there- 

 fore, that further proof of the para position of the side-chains in our 

 acid was necessary, and this we obtained by the oxidation of the acid 

 with potassic dichromate, according to the directions given by Beilstein 

 and Yssel. For this purpose 0.5 gr. of our acid was boiled in a flask 

 with a return condenser with four parts of potassic dichromate, and an 

 excess of sulphuric acid diluted with twice its volume of water, until 

 there was a copious white precij^itate ; this was then fdtered out, 

 boiled three times with water, and the residue converted into its dime- 

 thylester, which melted at 140-141°, thus proving tliat the product 

 was terephthalic acid, and that our acid was the paratoluylic acid. 

 We may add that a certain amount of terephthalic acid was always 

 obtained with the paratoluylic acid when we oxidized turmerol with 

 dilute nitric acid. 



Some unsuccessful attempts to make derivatives of the turmeryl- 

 chloride, which was described in a previous paper, J may be mentioned 

 here. The chloride was made by heating phosphorous trichloride 

 with turmerol for three hours. The product was purified by washing 

 its ethereal solution with dilute sodic hydrate and water, drying with 

 calcic chloride, and distilling off the ether. It formed a dark oil, with 

 a smell very different from that of turmerol, but, as it decomposed 

 even when distilled at a pressure of 12 mm., no attempt was made to 

 analyze it again. In the hope of obtaining an aldehyd or ketone 

 from it, 10 gr. of the chloride were boiled with an aqueous solution 

 of plumbic nitrate. Some plumbic chloride was formed, but the quan- 

 tity of chlorine removed was only 0.2363 gr., whereas the amount 

 calculated for the complete reaction was 1.868 gr. ; we obtained there- 

 fore only 12.65 per cent of the calculated amount, and, as the reaction 

 was so incomplete, it is not surprising that we were unable to isolate 

 the organic product. We also tried the action of potassic phthalimide 

 on the turmerylchloride in the hope of getting a turmerylamine, but 

 found there was no action at 100°, little or none at 150°. From 

 these results it is evident that turmerylchloride is not a very reactive 

 substance. 



* Ber. d. cli. G., XII. 615. J These Proceedings, XVIII. 11. 



t Ann. Chem., CCV. 132. 



