454 The American Naturalist. [May, 
of iron, they are encountered only in an accidental manner. 
The different modifications in the chemical composition of 
bones, depend essentially upon the nature of the filtrating 
water and by consequence with that of the strata which they 
percolate. 
It is the same as to the proportion of organic matters which 
diminish with time, but in a very irregular fashion according 
as the earth is more or less permeable. There is even to be 
found, sometimes, considerable organic matter in bones of 
great antiquity. The differences are too great between one 
deposit and another for us to be able in general to draw from 
the presence or the proportion of these elements, any induc- 
tion as to the length of time the bones have remained in 
the earth. 
The fixation of fluorine upon the phosphate of the bones is 
‘subordinate in a certain measure to the conditions of the de- 
posits and surrounding earth. The local circumstances have 
probably a much less influence because of the slowness of the 
phenomena. In any case, the series of analyses which are 
here given, show clearly that the proportion of fluorine in- 
creases at a perceptible rate during the later geological peri- 
ods, and that it can furnish in consequence better than the 
other elements a characteristic indication of the antiquity of 
the bone. 
The following conclusions seem to be justifiable. In the 
different deposits of the primary and secondary geologic 
epochs, the relative proportions of fluorine and of phosphoric 
acid are, upon the average, about the same as in crystallized 
apatite. In the tertiary and quarternary deposits there is a 
progressive and marked decrease in the proportion of fluorine, 
but this proportion remains during these epochs much higher 
than in modern times. It will, perhaps, be possible to use 
this means to fix the veritable age of certain human bones 
which have been found in the neighborhood of quaternary 
animals, but the deposits of which may have been disturbed 
_and the bones mixed. We cannot at present, from these ex- 
periments, establish this as a general method for the determi- 
nation, accurate or absolute, of the degree of antiquity of 
