— 115 — 
The residue of the evaporated urine was finally acidified with hydro- 
chloric acid, once more evaporated, and treated with alcohol and acetic ether. 
The fluorescence of the latter was but very faint. This time also no crystals 
could be obtained. 
Experiment No. 14. 2,0 g sodium anthranilate given subcutaneously to 
the same rabbit, caused no separation either of albumin or of sugar. The 
results of the examination of the urine were the same as in experiment No. 13. 
Experiment No. 15. A medium-sized dog was treated subcutaneously with 
2,0 g sodium anthranilate. No action. In the urine of 48 hours neither sugar 
nor albumin were present. The urine rendered alkaline with soda was evap- 
orated to a syrupy consistency, and tested for hippuric acid. There remained 
a residue which only dissolved in cold water when soda liquor was added. 
After the solution had been acidified with an excess of hydrochloric acid, the 
substance was again precipitated. The residue collected from the filter after 
filtration was repeatedly purified in this manner. The substance has an acid 
character, and forms salts with bases ; like hippuric acid, it dissolves with great 
difficulty in cold water and in ether, more readily in cold alcohol and acetic 
ether and hot water, and very rapidly in hot alcohol and hot acetic ether. 
The crystals of this acid wholly resemble those of hippuric acid, but it is not 
in this case a question of the last-named acid, whose melting point lies at 188°. 
The melting point of this acid lies between 315 and 318°. By means of 
potassium, potassium ferricyanide, and ferric chloride, the presence of nitrogen 
could be detected. 
Experiment No. 16. The same dog received an injection of 4,0 g sodium 
anthranilate. The urine of 48 hours, which showed a slight fluorescence, had 
an alkalme reaction, contained no albumin, and reduced a solution of copper 
already in the cold. As this reduction in the cold could doubtless only have 
been caused by sugar, it was remarkable that the fermentation-test had a 
negative result. For purposes of control two fermentation tubes were now 
prepared with a solution of sugar of equal percentage. To one of these 0,05 g 
sodium anthranilate was added, and so much tartaric acid, that the liquid 
had a shghtly acid reaction. After 6 hours the fermentation-process had in 
the one tube come to an end, whilst in the other tube, to which the sub- 
stance had been added, no carbon dioxide had as yet evolved. Only after 
24 hours a quantity of carbon dioxide had been formed which was less than 
half that in the other tube. Further fermentation did not, however, take 
place in the following 24 hours. Anthranilic acid or its salt, consequently, acts 
on the yeast, and retards, and therefore curtails, the process of fermentation 
considerably. 
The presence of sugar in the urine was definitely proved by the phenyl- 
glucosazone test. Many and well-developed crystals of the characteristic form 
were obtained. 
From the presence of sugar, and the negative result of the fermentation 
test, it may be concluded that the urine contained anthranilic acid, and that in 
such abundance, that it was capable of acting on the yeast. 
The urine was now treated in the same manner as in experiment No. 13. 
The alcohol, and also the acetic ether, showed a beautiful blue fluorescence. 
But the acid which I obtained in experiment No. 15, could not be produced. 
Experiment No. 17. A small dog, weighing about 16 pounds, was sub- 
cutaneously injected with 5,0 g sodium anthranilate. After 24 hours the 
animal was attacked by diarrhoea, and on the second day it took absolutely 
no food. On the third day it had completely recovered. 
The urine collected in 48 hours was alkaline, showed a blue fluorescence, 
contained no albumin, and reduced solution of copper in the cold. The fer- 
mentation-test was negative; the phenylglucosazone-test yielded positive results. 
8* 
