32 BOTANICAL GAZETTE [JANUARY - 
nine days, however, pigment makes its appearance. In sucha 
nutrient medium, then, there is a struggle between two of the 
vital activities of this organism, namely, its ability to produce 
acid and its power to form alkali. The latter ‘ function”’ in this 
case ultimately gains the upper hand. 
In 3 per cent. saccharose broth the course of events is exactly 
reversed. Saccharose is less easily fermented by this species 
and the result is that pigment is formed at the outset and a two 
days’ growth is well colored. As the saccharose becomes con- 
verted into acid the color slowly fades out and at the end of 
four days it has vanished. 
The pigment formed by all the species is at first, as has been 
shown by Thumm and others, a delicate robin’s egg blue, but as 
the solution becomes more alkaline owing to the bacterial growth, 
the color changes to green, and in strongly alkaline solutions is 
a deep green tint, showing no fluorescence. The addition of 
alkali to a solution containing the blue pigment produces at once 
the same change as is wrought more slowly by the alkali formed 
by the bacteria. 
SUMMARY. 
The upshot of my experiments may be summarized under the 
following heads: 
1. Lhe presence of both phosphorus and sulfur is essential to the 
Jormation of the fluorescent pigment. | 
The effect of almost infinitesimal quantities of sulfate in the 
presence of phosphate compels us to accept with some reserve 
the statements made by authors as to the production of fluores- 
cence in media devoid of sulfur. Thumm’s curious statement™ — 
that B. 72. albus, while it produces fluorescence in a medium com- 
posed of I per cent. ammonium succinate, 0.1 per cent. potassium 
phosphate, 0.04 per cent. magnesium sulfate, 0.02 per cent. cal- 
cium chlorid, and also produces a feeble green fluorescence when 
the magnesium sulfate is omitted, produces no fluorescence at 
all if the calcium chlorid also be left out, is perhaps most easily 
** Op. cit. 42-43. 
