87 
being so. These statements seem to have many exceptions; 
as, for example, the colored transparent salts of metals, show 
the same color by reflected as by transmitted light. But I 
am persuaded that a careful discussion of the case would 
show that such exceptions are not well taken, and that this 
apparent discrepancy with the statements may be consistently 
explained away; thus it may be shown that the supposed 
reflected light of the exceptions is really a part of the 
transmitted light which has been returned by internal 
reflection; such mixture of the transmitted with the reflected 
light, implies a considerable degree of transparency of the 
substance under test. The lustre and whiteness of metals 
have a close relation to their opacity and density; perhaps 
the relation is that of effect and cause. If the opacity and 
density of metals be progressively decreased the optical 
metallic character will in the same ratio be diminished; the 
true color by reflected light would become brighter and freer 
from white light till it come to be contaminated with more 
and more of the returned transmitted light. Such changes 
are beautifully exemplified by the gradual additions of a 
solvent to Fuchsine or other Aniline colors in crystals. 
Aniline colors, Prussian Blue, Indigo, Carmine and all other 
dye-stuffs which have very great tinctorial powers, have the 
metallic lustre, and their color by transmitted light is nearly 
complementary to that by reflected light. 
In their relation to light I suggest that metals are closely 
analogous to those dye-stuffs which show a bronzed surface 
by reflected light. Metals are more perfectly bronzed 
because their opacity and density are greater, or in other 
words their tinctorial powers are greater. 
It will be seen that the above theory requires for. its 
demonstration a transparent diluent or solvent of metals, 
which shall have no chemical action.on them. Such a 
solvent, for a few of the metals, is anhydrous liquid Ammonia. 
If this menstrum be gradually added to the silver-white 
alkali metals, the whiteness disappears, is replaced by copper- 
redness which at last gives place to the blue of transmitted 
light. The changes of tint in this case from copper-redn ess 
