258 ANNUAL EEPOET SMITHSONIAlSr INSTITUTION, 1919. 
violet tint in the color. If, liowever, using any one of the three 
alkalies, a highly acid glass containing the same amount of cobalt 
oxide be made, the difference between it and the comparable soft 
glass is very marked. There is much less depth of color altogether. 
What there is is a somewhat violet blue in the case of the potash 
glass, a rather lighter and more violet blue in the soda glass, and a 
still lighter pink violet in the lithia glass. This nearly pink glass 
goes to a weak but quite distinct blue when heated. 
The effect of the alkalies potash and soda on the color of borates 
is not so marked in the case of cobalt as it is in that of nickel. 
Lithia in comparison with them always gives for equivalent pro- 
portions a much more decided violet tint in the blue. The influence 
of the proportion of base to acid, however, is marked, and can be ver}^ 
well seen by using any one of the alkalies in varying proportions wdth 
molten boric anhydride, to which a small amount of cobalt oxide 
(about 0.25 per cent) has been added. [Cobalt oxide dissolves in 
highly heated boric anhydride, but on cooling cobalt borate separates 
out, giving an opaque, very pale blue, glassy mass.] Taking lithium 
carbonate as the alkali and adding only a very small amount (about 
0.25 per cent) the whole of the cobalt remains in solution on cooling, 
and the resulting glass is seen to be blue while still hot, to change to a 
more and more violet tint on cooling, and to be almost a pink when 
cold. The addition of more of the alkali intensifies the blue color, 
giving greater depth of color, and the mass Avhen cold is a violet 
blue. Similar variations in color can be obtained with equivalent 
quantities of potash and soda, but the effect of these alkalies is 
alwaj'S to give a more pronounced blue as the amount of alkali is 
increased. Comparing the weakest base, lithia, then, with potash, 
the strongest, and progressively adding each to borate beads con- 
taining cobalt, no amount of lithia, up to the point when it is im- 
possible to keep the bead vitreous, will giA^e as blue and as strongly 
colored a bead as the equivalent, or even less than the equivalent, of 
potash will produce. There is ahvays a more violet tint in the lithia 
bead. 
In conclusion, a brief reference may be made to another coloring 
agent, copper oxide. This oxide is not soluble in boric anhydride 
when a bead of the latter is heated in an oxidizing flame, but by the 
addition of an alkali a clear blue bead is obtained. Should the alkali, 
e. g., potash, be added in very small amount, so as to give a highly 
acid mixture of about the composition, for instance, represented by 
the portions K2O.5OB2O3, the coloration due to about 0.25 per cent of 
copper oxide is so faint that the bead is practically colorless, although 
this amount of the oxide is sufficient to give a markedly blue bead 
in potassium bi-borate, K2O.2B2O3. In this case, also, then the color 
becomes more intense as the amount of alkali used is increased. 
