JACKSON AND BOOS, — COLORED COMPOUNDS. 177 



we next studied the eiFect of heat upon it. A sample dried for twenty- 

 four hours in vacuo was heated to 100°-llO°. After one hour's heating 

 it had lost 0.0002 gram, and upon further heating remained constant. 

 The heat was then increased to 130°-140°, but even at this temperature 

 there was no loss in weight. In view of this experiment it seems im- 

 possible to maintain that the substance contains methyl alcohol of crys- 

 tallization. When the temperature was raised to 165°, the substance 

 decomposed, as was shown by a change in color. 



Properties of the Addition Product of Trinitranisol and Sodic Methylate. 

 — It crystallizes from methyl alcohol (with slight decomposition) in prisms 

 apparently belonging to the monoclinic system, of a beautiful scarlet, or 

 perhaps rather vermilion color with a bronze reflex. It decomposes at 

 about 165°, and, if heated in the free flame, explodes with great violence. 

 In the dry state it is comparatively stable, keeping well even when 

 exposed to the air ; but in contact with water it is gradually decomposed, 

 forming trinitranisol, which, if the action is very long continued, passes 

 into picric acid. Methyl alcohol dissolves it, but not without decom- 

 position, as some trinitranisol is found mixed with the crystals obtained 

 from such a solution ; it is almost insoluble in benzol. Acids decompose 

 it at once, forming a body, which after recrystallization from alcohol 

 melted at 64° and was therefore evidently trinitranisol ; methyl alcohol 

 was also set free, as shown by the odor. 



Action of Ethyl Alcohol on the Addition Product of Trinitranisol and 



Sodic Methylate. 



This work was undertaken in the hope of deciding whether the addi- 

 tion product contained methyl alcohol of crystallization, as in that case 

 the methyl alcohol must be replaced by ethyl alcohol upon recrystalliza- 

 tion from this solvent, whereas, if the second methyl group was not pres- 

 ent in alcohol of crystallization, it might either remain in the compound 

 under these conditions, or be replaced by the ethyl group. If, therefore, 

 the methyl compound were unaltered by crystallization from ethyl alco- 

 hol, it could not contain methyl alcohol of crystallization. Unfortunately, 

 the experiment showed that the methyl group was replaced by ethyl, and 

 therefore no inference in regard to the constitution of the body could be 

 drawu from this work. 



On recrystallizing the methyl compound from ethyl alcohol a sub- 

 stance was obtained closely resembling the original methyl compound 

 in color, crystalline form, and properties, but which gave the following 

 result on analysis after having been dried in vacuo : — 

 VOL. xxxni. — 12 



