283 



of several days. Admixed with much triple phosphate, as shown 

 first by the microscope, and afterwards by the chemical analysis. 



Bibasic phosphate of magnesia 4 '30 



Bibasic phosphate of lime 5'4 1 



971 



Fourth Sample. Separated from six ounces of fresh urine. De- 

 posit very pure. 



Bibasic phosphate of lime 1 '96 



No phosphate of magnesia. 



Now the admixture of the phosphate of magnesia in the first three 

 samples was due solely to the fact that the phosphate of lime, 

 deposited at first in the pure state, was allowed to remain in the 

 urine until decomposition had commenced, and the phosphate of 

 magnesia and ammonia had, in consequence, become formed. De- 

 posits of phosphate of lime are sometimes contaminated from the 

 same cause with carbonate and oxalate of lime. 



These analyses are therefore conclusive as to the composition of 

 this earthy phosphate. In order to show that no error has been 

 committed in them, I here append the process adopted. That the 

 deposit in question really consisted of a phosphate, was first repeatedly 

 determined by the action of a solution of nitrate of silver ; the 

 crystals when touched with this reagent assumed a bright golden 

 yellow colour. After having been separated and washed in distilled 

 water, the phosphate was ignited to free it from animal matter, urea, 

 &c., and weighed. It was then dissolved in hydrochloric acid; 

 ammonia was added until a permanent precipitate formed ; this was 

 redissolved by the addition of acetic acid. First the lime was preci- 

 pitated from the solution by oxalate of ammonia, and afterwards the 

 magnesia as follows : chloride of ammonium was added, then ammonia 

 in slight excess, and lastly, phosphate of soda. The oxalate of lime 

 formed was converted into carbonate of lime in the ordinary manner, 

 and the phosphate of ammonia and magnesia into the pyro-phosphate 

 of magnesia ; these were then weighed separately, and the amounts 

 of the bibasic phosphate of lime were determined by the usual calcu- 

 lations. The results obtained corresponded very closely with the 

 original weights of the ignited phosphates subjected to analysis. The 



VOL. X. X 



