1895.] Fluorine as a Test for the Fossilization of Animal Bones. 311. 
all the phosphoric acid is found in the acetic solution with 
which we proceed to a volumetric test. It is unadvisable to 
pour into the acetic solution the standard solution of ura- 
nium, because the latter will cause some phosphate of lime 
to precipitate with the phosphate of uranium and we should, 
in consequence, find too small a proportion by measuring the 
volume of uranium solution employed. 
We arrive at a more exact result by making the determina- 
tion by the indirect method: We dilute the liquid to 200 
cubic centimetres, then pour it, with the aid of a graduated 
burette, into a measured volume of a standard solution of 
nitrate of uranium, to which is added a little acetate of soda 
‘slightly acidified by acetic acid and the solution is heated to 
near 80 degrees. We stop when one drop of the liquid, placed 
on a porcelain saucer, in contact with a little ferrocyanide of 
potassium in powder, will cease to give a brown color. In order 
to standardize the solution of uranium, we operate in the same 
manner, starting with an exactly known weight of phosphate 
of lime. : By taking phosphate of lime exactly analyzed, we 
succeed better than by taking pure phosphate of soda, because ~ 
we standardize and make the determination in conditions 
which are identical. We add pure water to such extent that 
40 cubic centimetres correspond exactly to 0.200gr. of phos- 
phorie acid (or 1 cubic centimetre to 0,005gr. of P,O,). The 
operation should be performed a second time for verification. 
In the assay of a phosphate, if one operates on 2 grams and 
makes the acetic solution up to 200 cubic centimetres, we can 
calculate the proportion of phosphoric acid from the volume 
V of thefliquid which will have saturated the 40 cubic centi- 
metres of the solution of uranium. The volume contained 
0.200gr. of phosphoric acid. As a consequence, in 1 gram of 
phosphate, or 100 cubic centimetres of the liquid, there would 
be Ogr, 200+" (V being expressed in cubic centimetres). 
The following table gives the results of analyses for a certain 
number of bones of varied origin. 
These analyses confirm at nearly all points the conclusions 
that M. Fremy has drawn from his numerous analyses. We 
remark with him as to the constancy in the proportions of 
