( 282 ) 



that, in the many cases which I have investigated, the presence of 

 17o of an isomer causes about 0.5° depression in the melting point, 

 the fact that the oil did not solidify till considerably below 0° and 

 again melted at a slight elevation of temperature whilst the pure 

 isomers did not liquefy till 60° or above, cannot be explained by 

 the presence of relatively small quantities of 1,3,4-dinitrotoluene in 

 presence of one other constituent, but it must be supposed to consist 

 of a ternary system. This was verified when the fractionation was 

 continued still further; soon, the fractions with the lowest boiling 

 points began to solidify on cooling, or slowly even at the ordinary 

 temperature, and the solidified substance proved to be 1,2,3 — dini- 

 trotoluene. The fractionation combined with the freezing of the 

 different fractions then caused the isolation of a third isomer namely 

 1,3,6 — dinitrotoluene, so that the three isomeric dinitrotoluenes 

 desired had thus all been obtained by the nitration of 7?i-nitroto- 

 luene. The fourth possible isomer (1,3,5) could not be observed even 

 after continued fractionation and freezing. 



As regards the relative quantities in which the three isomers, 

 detected in the nitration product, are formed, it may be mentioned 

 that this product consists of more than one half of 1, 3, 4-dinitro- 

 toluene, whilst (1, 2, 3) seems to occur in larger quantities than 

 (1, 3, 6), as the isolation of the latter in sufficient quantity gave the 

 most trouble. 



The corrected solidifying points of the dinitrotoluenes (the sixth, 

 symmetric one was prepared by Beilstein's method A. 158, 341 in 

 order to complete the series) were determined as follows. Those of 

 the dinitrobenzoic acids and of their ethyl esters are also included 

 in the subjoined table. 



The specific gravities of the dinitrotoluenes and the ethyl esters 

 were determined by means of Eykman's picnometer at 111°.0 with 

 the following result : 



