256 ANNUAL REPOET SMITHSONIAN INSTITUTION, 1919. 
remarked that potash glasses generally give a good violet and soda 
glasses a brown or a brownish violet. Using small amounts of 
manganese dioxide in batch mixtures, as free as possible from iron, 
it has been found possible to get glasses practically colorless to the 
eye, some of which readily give color on exposure to radium for a 
period during which others develop no color. In making the latter, 
the conditions in the furnace were arranged for complete reduction 
of the manganese dioxide. In making the former, as little deoxida- 
tion as possible was aimed at. In one instance thin rods drawn from 
the melt of one of these, in which very little manganese dioxide was 
used, cooled almost colorless, but "struck" quite a marked violet 
color on reheating. This chilled glass was also very sensitive to 
radium. More urgent work prevented further experiments, but the 
facts so far obtained are mentioned as relevant to the question of the 
chemical condition of coloring agents in glasses, and as an illustra- 
tion of one which would appear to be somewhat of a border line ex- 
ample of the groups (a) and (6), referred to previously. I am 
reluctant to dismiss the matter in this rather summary fashion, but 
the interesting speculations which will occur to many can hardly be 
dealt with shortly. I would, however, recall the well-known pink 
or violet color to be seen in some window glasses which have been 
exposed for years to daylight. In all examples which I have been 
able to examine, manganese has been found to be present, and I can 
imagine that the color has developed in daylight in a manner similar 
to that in which it has been found to be developable in manganese 
glasses by radium, by cathode discharge, or by heat. The color of the 
old window glasses is a little puzzling, if they are soda-lime glasses. 
One would expect them to be broAvner in tint; but perhaps on in- 
sufficient grounds, since no direct experiments have, so far as I know, 
been made with manganese glasses made with potash and with soda- 
batch mixtures and exposed to sunlight for a long period. It is a 
matter for regret that when the old tinted window glasses were ex- 
amined for manganese the idea of the influence on the color of the 
alkalies present did not occur. [Having regard to the effect of 
manganese greatly to enhance the phosphorescence of potash and 
soda-lime glasses, and to the known coloration of certain potassium 
and sodium compounds under cathode discharge, it is possible that 
the colors in the old window glasses described are not due to man- 
ganese dioxide itself, but that manganese may have rendered the 
alkali compounds in the glasses more sensitive to light of short wave 
lengths. The fact that glasses containing no manganese did not color 
under cathode discharge, etc., is not conclusive, since such glasses 
showed but feeble phosphorescence. The observation, however, that 
of two glasses containing the same amount of manganese, and giving 
equal phosphorescence, the one in which there is evidence of some 
