OP ARTS AND SCIENCES. 313 



and XXIX., in which the mixture stood one or more days longer 

 than the eighteen hours used for the other experiments. In fact, 

 longer standing could not have any great effect in the case of the 

 ethylate, as in eighteen hours eighty to ninety per cent of the tri- 

 bromtrinitrobenzol had entered into the reactions. With the higher 

 alcoholates this was not the case, and it may he that the action had 

 not reached its end in eighteen hours, and that the variation ob- 

 served in the percentages was due to differences in the times of re- 

 action; but some experiments, in which the mixtures were allowed 

 to stand a longer time in order to test this point, gave unsatisfac- 

 tory results, apparently on account of the oxidation of the sodic 

 nitrite formed by the oxygen of the air. 



PART III. 



Action of Some Ethylate on Certain" Derivatives op TRI- 

 BROMTRINITROBENZOL CONTAINING AlKYLOXY RADICALS. 



Action of Sodic Ethylate on the Triphenylether of Trinitro- 

 phloroglucine. 



This work was undertaken in the hope that the triphenylether 

 of trinitrophloroglucine, C 6 (OC 6 H 5 )3(N02)3, might behave toward 

 sodic ethylate like tribromtrinitrobenzol, that is, lose one or more 

 nitro groups as sodic nitrite, which would be replaced by the ethoxy 

 group. This hope, however, has not been fulfilled, as instead of the 

 nitro groups the phenoxy radicals are removed by the sodic ethylate, 

 becoming replaced by ethoxy groups, while sodic phenylate is 

 formed. The experiments were carried on as follows. 2.5 gr. of 

 the triphenylether of trinitrophloroglucine* dissolved in benzol 

 were mixed with the sodic ethylate from 0.4 gr. of sodium, and the 

 bright red mixture allowed to stand two or more hours in the cold. 

 After this it was allowed to evaporate spontaneously, and the resi- 

 due warmed with water on the steam bath, when most of it dissolved. 

 The insoluble portion was purified by crystallization from alcohol 



* In preparing this substance from tribromtrinitrobenzol again we have 

 found that a tolerable excess of sodic phenylate should be used in order to avoid 

 the formation of a compound containing bromine, — C(-(OC 6 H 5 ) 2 Br(N(>,)3 

 probably. With such an excess, the reaction runs without heat, and is complete 

 in a few minutes. We have not observed the green color noticed by us last 

 year (these Proceedings, XXV. 188), which must therefore have been due to 

 some impurity. 



