— 114 — 
it was then again filtered, and the filtrate exactly neutralised with sulphuric 
acid. After evaporating this neutral filtrate, it was repeatedly extracted with 
alcohol, the alcohol again driven off, and the small residue acidified with acetic 
acid, and precipitated with acetate of copper. In order to remove the copper 
from the liquid thus obtained, it was treated with sulphuretted hydrogen, and 
filtered. The filtrate had a faint yellowish colour, with a slight blue fluores- 
cence. The melting point could not be determined, as the quantity remaining 
after slowly evaporating the water was too small. But the fluorescence could 
only have been caused by anthranilic acid. 
If the fatal dose of anthranihc acid in frogs is not less than 3,0 g per 
kilogram, a more powerful toxic effect of the substance on warm-blooded anim- 
als may hardly be expected. 
Experiment No. 12. A rabbit was injected with 1,0 g sodium anthranilate, 
without occurrence of any symptoms whatever. The quantity of urine obtained 
in 24 hours was 30 cc; it had an alkaline reaction, and contained neither 
sugar nor albumin. No aniline could be detected in the urine by means of 
chloride of lime. The urine was next added with alkali in order to test 
it for hippuric acid; it was evaporated to the consistency of syrup, the residue 
repeatedly extracted with cold alcohol, the alcoholic extract evaporated, de- 
composed with hydrochloric acid, and repeatedly shaken with acetic ether. In 
order to isolate benzoic acid, which possibly might be present, the residue, after 
driving off the acetic ether, was extracted with petroleum ether. But not even 
traces of benzoic acid could be detected. The residue from the acetic ether 
was dissolved in alcohol, when immediately a beautiful blue fluorescence 
appeared, which could only be due to the presence of anthranilic acid. One 
half of the solution was poured off, the alcohol evaporated, and the residue 
taken up with water. When potassium-bismuth iodide was added, a yellow-red 
precipitate was formed, but the pure anthranilic acid could not again be obtained. 
In the remaining half of the alcoholic solution some crystals floated at the 
bottom of the vessel, which could only be dissolved with difficulty in cold alcohol, 
and not at all in ether. By means of ether they could very readily be freed from 
the greasy matter which still adhered to them. In hot water the crystals dis- 
solved. After slowly evaporating the water, there could be detected with the 
microscope , firstly small rhombic prisms , secondly crystals arranged in druse- 
like form. The yield was too small for making any determinations whatever. 
Anorganic constituents were not present in the crystals. 
Experiment No. 13. Even 5,0 g sodium anthranilate, injected subcutane- 
ously in a medium-sized rabbit, had no effect. The urine collected in 48 hours 
had an alkaline reaction, and did not reduce a solution of copper with the 
application of heat. The fermentation test gave negative results. In about 
15 cc urine submitted to the phenylglucosazone test, the precipitate contained 
only a few star-shaped crystals, quite isolated and badly formed, although the 
urine had been cooled very slowly on a water bath. They, however, proved 
that the urine contained traces of sugar. 
Small traces of albumin could also be detected by means of Esbach's 
reagent, and the underlaying test with nitric acid. 
The urine was rendered still somewhat more alkaline with soda, evaporated 
to the consistency of syrup, repeatedly extracted with alcohol, the alcohol ex- 
tract evaporated, the residue decomposed with hydrochloric acid, and repeatedly 
shaken with acetic ether. The acetic ether showed intense fluorescence. As the 
fluorescence sufficed for the detection of anthranilic acid, the production of 
the acid in the pure state was not proceeded with. After driving off the 
acetic ether, the residue was taken up with water, rendered alkaline with 
soda, filtered, and acidified with an excess of hydrochloric acid. But no crystals 
as in the previous experiment could be obtained. 
