156 PROCEEDINGS OP THE AMERICAN ACADEMY 



Although we have not found it possible to separate these acids 

 directly by fractional crystallization, they may be separated by means 

 of their calcium salts. While we have found it easy to prepare in 

 this way small quantities of the pure acids, the complete separation 

 of large quantities of the acids involves a very considerable amount of 

 time and labor. We have found it most convenient to dissolve the 

 mixed acids in ammonic hydrate, and to ad I calcic chloride to the 

 dilute solution (1 : 30). A very sparingly soluble crystalline calcium 

 salt is thus thrown down, which contains chiefly one of the two dibrom- 

 pyromucic acids, while the calcium salts of the other acids remain in 

 solution. The sparingly soluble calcium salt may be purified by re- 

 crystallization from boiling water, or the acid may be set free, the 

 calcium salt reprecipitated, and the operation repeated until the melting- 

 point of the acid is found to be 164-16G°, The solution containing 

 the more soluble calcium salts is acidified, and the acids thus obtained 

 are separated by recrystallization from boiling water. The tribrom- 

 pyromucic acid is almost insoluble in water, while the 8-monobrom- 

 pyromucic acid, though but slightly more soluble in cold water than 

 tlie dibrompyromucic acid, is present in such small quantity that its 

 removal is a matter of little difficulty. When the melting-point of 

 the dibrompyromucic acid has been raised to 184° by recrystallization 

 from water, its further purification is effected by washing or recrystal- 

 lizing from benzol. The acid whose calcium salt is more sparingly 

 soluble in water, and which, as we shall afterward show, is the (38- 

 dibrompyromucic acid, melts at 1G7-168°, while the isomeric ^y-di- 

 brompyromucic acid, whose calcium salt is more readily soluble, melts 

 at 191-192°. Although we have made many attempts to control at 

 will the quantities of the two dibrompyromucic acids formed, we have 

 as yet met with no success. The yield of the mixed acids was far 

 from satisfactory, and amounted to but from thirty-four to thirty-eight 

 per cent of that theoretically required. Tiie difficulties in the way of 

 complete separation of the two acids make it impossible for us to give 

 the precise ratio between the amounts of the two isomers found. While 

 we have usually obtained nearly twice as much of the low-melting acid 

 as of the high, we are inclined to think that the original mixture 

 contained them in nearly equal quantities. 



The alcoholic filtrate from the sodium snlts contained, beside small 

 quantities of these salts in solution, considerable quantities of sodic 

 8-monobrompyromucate. When diluted with water, it became turbid, 

 with the separation of a small quantity of a colorless oil. In order 

 to determine its nature, we distilled the filtrate with steam and precipi- 



