OP ARTS AND SCIENCES. 141 



low a temperature as 175°. It is but slightly soluble in alcohol, even 

 when boiling, essentially insoluble in it when cold, soluble in chloro- 

 form, and more easily in ether, benzol, acetone, glacial acetic acid, or 

 carbonic disulphide. Chloroform, or a mixture of benzol and alcohol, 

 is the best solvent for it. 



The tribromtriuitrobenzol is a decidedly reactive substance, forming 

 compounds with most of the common reagents ; of these compounds 

 we have been able to study thoroughly only those derived from alco- 

 holic ammonia and aniline, which will be described later in the paper, 

 but some preliminary experiments with other reagents may find a 

 place here. With potassic hydrate, dissolved in alcohol, a yellow 

 product was formed which gave red potassium and yellow barium 

 salts, the latter being only slightly soluble, and separating in hair-like 

 crystals from its hot aqueous solution. It is probable that this pro- 

 duct is the trinitrophloroglucin of Benedikt * but to decide this point 

 the experiment must be repeated with a larger quantity of substance. 

 With sodic ethylate it gives what appears to be a new compound. 

 When boiled with an alcoholic solution of potassic sulphocyanate, it 

 forms a dark red powder, which we have not yet succeeded in obtain- 

 ing in crystals. When heated in a sealed tube with potassic iodide 

 and alcohol to 150° for 18 hours, a crystalline compound is formed, 

 which has a very high melting point and explodes when heated to a 

 somewhat higher temperature ; but the yield is so small that we have 

 not been able as yet to obtain enough of it sufficiently pure for analy- 

 sis. All these substances will be more thoroughly studied in this 

 laboratory, and the behavior of tribromtriuitrobenzol with other re- 

 agents, especially sodium malonic ester, will be investigated also. 

 When tribromtriuitrobenzol is heated to 100° with common strong 

 sulphuric acid, it dissolves, but crystallizes out unaltered as the solu- 

 tion cools. A boiling solution of argentic nitrate in alcohol has no 

 action on it, and the same remark applies to argentic nitrite, as was to 

 be expected. We hope, however, that the triiodtrinitrobenzol may 

 react with this latter substance, and it was for this reason that we 

 undertook the study of the action of potassic iodide on the tribromtri- 

 nitrobenzol. 



In the hope of obtaining addition-products similar to those formed 

 by Hepp's trinitrobenzol with hydrocarbons, we have studied the ac- 

 tion of tribromtrinitrobenzol on naphthaline. For this purpose benzol 

 solutions of the two substances were mixed in the proportion of one 



* Ber. d. ch. G., xi. 1376. 



