CRYSTAL RIVER REVISITED. 421 
Several masses of coral, hemispherical in shape, lay in the sand or in the shell 
layer, the flat surface of each mass showing marks of use as a smoothing-stone. 
In sand thrown out at the time of the former investigation was part of an 
ear-ornament of sheet-copper, of the form sometimes designated “ spool-shaped." 
The under portion of the ornament is somewhat defective by reason of former 
breakage; the upper part is entire. 
On the upper surface of this upper part is a thin sheet of material resembling 
iron rust. That this material was not deposited after the ornament was placed in 
the mound (and we know that bog-iron sometimes is deposited in this way) is 
shown by two facts: The material does not lie over the entire ornament, but is 
symmetrically placed on the upper, outer surface, just as silver plating has been 
found upon similar aboriginal ornaments; and, secondly, an even covering of 
decayed wood is still present above the material. Evidently an even deposit of 
bog-iron would not form on one part of an ornament alone, and especially below a 
portion protected by a covering of wood. Therefore the plating was artificial. 
As the ores of iron are not malleable and, therefore afford no material from 
which a coating of the kind found by us could have been directly made, the original 
plating must have been of metallic iron. 
Small fragments of this coating, analyzed by Dr. H. F. Keller, proved it to be 
hydrated oxide of iron, containing nickel. This hydrated oxide could well be a 
derivative of metallic iron which had rusted through and through, and this hypoth- 
esis is strengthened by the fact that the material on the ornament is magnetic, 
while hydrated oxide ores of iron are non-magnetic, as a rule. 
As it is practically certain that at Crystal River we have to do with a pre- 
Columbian site, we must reckon with the fact that this sheet-iron was of aboriginal 
origin solely. 
It was not in the power of the aborigines to recover the metal iron from iron 
ore, and terrestrial native iron is not found in various localities in masses as the 
native, metallic copper is found; consequently, it is necessary to explain how pre- 
Columbian aborigines in Florida became possessed of iron in its native, metallic 
form. 
Fortunately this can be done with the aid of Prof. F. W. Putnam!’ and Dr. L. 
P. Kinnicutt,? who have shown that iron from the Turner group of mounds in Ohio 
is meteoric iron, containing a large percentage of nickel, and showing other distinc- 
tive features. 
In the plating of our specimen, which is rusted through and through, and 
from which only minute fragments could be taken, exact determination of compo- 
nent parts could not be made, but Doctor Keller has noted the presence of nickel? 
! Sixteenth Report, Peabody Museum, Cambridge, 1882. 
2 Seventeenth Report, Peabody Museum, Cambridge, 1883. | 
3 Doctor Keller found but a small quantity of nickel in our specimen, while the amount of nickel 
in the Ohio nuggets is relatively large. The analysis of the iron from Ohio was confined to nuggets of 
metallic iron. Doctor Keller thinks much nickel must have washed from the thin plating of our 
specimen during the transformation from metallic iron into iron rust. 
