OF ARTS AND SCIENCES. 203 



in long needles, which were frequently collected in globular aggre 

 gatioDs. The salt was hardly more soluble in hot than in cold 

 water, and could best be obtained by the evaporation of its cold 

 aqueous solution in vacuo over sulphuric acid. In this respect it 

 closely resembled the baric /3-sulphopyromucate. An anatysis 

 showed that the salt was baric sulphomethylpyromucate, and that it 

 contained five molecules of water of crystallization. 



I. 0.4798 grm. air-dried salt gave 0.2597 grm. BaS0 4 . 

 II. 0.6078 grm. air-dried salt lost at 165° 0.1383 grm. H 2 0. 



Calculated for Found. 



BaC c H 4 S0 6 5H 2 0. I. II. 



Ba 31.78 31.83 



H 2 20.87 20.71 



0.5272 grm. salt dried at 165° gave 0.3598 grm. BaS0 4 . 



Calculated for Ba C 6 H 4 S0 6 . Found. 



Ba 40.17 40.13 



Bromine water gave in an aqueous solution of the salt no precipi- 

 tate even on boiling, but after this treatment with bromine, if baric 

 hydrate was added to alkaline reaction a white flocculent precipitate, 

 which closely resembled baric sulphofumarate, appeared. This be- 

 havior showed conclusively that the sulpho group had not entered 

 the furfuran ring in the 8 position, and the conclusion seems war- 

 ranted that it had not done so only because the methyl group 

 already occupied that place. 



Brommethylpyromucic Acids. 



Bromine acts readily upon methylpyromucic acid at ordinary 

 temperatures, and substitution is so easily effected that the isola- 

 tion of an addition product is a matter of considerable difficulty. 

 The instability of the addition product renders it unsuitable for 

 the preparation of substituted acids, and facilitates the formation of 

 satisfactory products by direct substitution. The nature of the 

 products thus formed is, however, largely dependent upon the tem- 

 perature at which the reaction takes place. At low temperatures 

 the bromine enters first the furfuran ring and afterwards the methyl 

 group, while at higher temperatures the methyl group is first attacked, 

 altbough substitution apparently takes place at the same time to a 

 certain extent in the ring. The facts which we have here observed 

 evidently correspond precisely with those so thoroughly established 



