Dr. E. J. Mills on Nitrated Toluol 19 



1 therm.) • in crystalline form it was much more prismatic and 

 less platy than either of the preceding. 



Experiments were made as to the solubility in absolute alcohol 

 of F,, F 2 , K, and M. One hundred parts of the solvent took up 

 1-540 per cent, of ¥ 1 at 15°*3, 1*557 of F 2 at 15°'7, 1*615 of K 

 at 15°*1, and 1-427 of M at 15°*1. Hence the error of the 

 process (F l being regarded as identical with F 2 ) is less than '017 

 on the above numbers, and K and M are distinct from each 

 other and from F. 1 could hardly find any reliable difference 

 between the solubilities of F 2 and K at about 4°*75; but K still 

 gave the heavier residue. At 9°*6 the solubility of F } was 1*204 

 per cent. • of M, 1*097 per cent, at 9°*3. These numbers are 

 confirmatory of those in the first group. 



Remarks. 



Trinitrotoluol. — Of the three modifications above referred to, 

 and all of which are obtained by direct processes, the one pre- 

 pared from liquid mononitrotoluol requires the least expenditure 

 of work, and has the greatest solubility. Its fusion-point is 

 nearly identical with that prepared, though with more difficulty, 

 from toluol itself; but this second modification shows a dimi- 

 nished solubility. The third modification is made from solid 

 mononitrotoluol, with the greatest expenditure of energy ; it has 

 an unmistakably higher fusion-point, but a lower solubility than 

 either of the two preceding isomers. 



Dinitrotoluol. — Solid mononitrotoluol is more difficult to con- 

 vert into dinitrotoluol than either toluol or liquid mononitrotoluol, 

 and melts at a distinctly higher temperature than either of them 

 as they are when freshly prepared. Having regard, however, to 

 the shifting fusion-point of the product from liquid nitro toluol, 

 the difficultly observed fusion-point of the product from toluol 

 itself, and the impracticability of determining the solubility of 

 either of these isomers in alcohol, I cannot quite decide whether 

 the toluol- and liquid nitrotoluol-products are distinct. Judging 

 from their obvious mimicry of the trinitrotoluols, I should 

 presume that they are. 



Mononitrotoluol.— -The facile process of separating the liquid 

 from the solid isomer, referred to in preceding paragraphs, will, 

 it is hoped, save many weary fractionations. That the solid 

 modification is produced in larger quantity when more energy is 

 employed in the reaction, is already known as a result of Rosen- 

 stiehFs labours. 



The fusion-points I have given in this paper are good in the 

 second decimal place. That, however, will probably have to be 

 lowered by a unit or two, on account of the difference between 

 Kopp's correction and mine for the thermometers' exposure; 



C2 



