462 BOTANICAL GAZETTE [DECEMBER 
pressure valves so that a small stream of gas at normal pressure 
ran constantly through the germinators. 
The gas was washed by passing it successively through potash 
bulbs containing concentrated potassium permanganate solution, 
and 33 per cent potassium hydroxide. The whole series of coils 
and jars was packed in ice for 12 hours after the hydrogen began 
to flow, and the gas was carefully analyzed by phosphorus absorp- 
tion until it was found to be coming from the apparatus with as low 
oxygen content as when taken directly from the tank, at which 
time the temperature was allowed to rise sufficiently for germina- 
tion. In this way the possibility of the initiation of germination in 
only a partially replaced atmosphere was precluded. The appara- 
tus was quickly brought to ordinary temperature, and kept con- 
stant at 21.5° C. by allowing a current of water from Lake Michigan 
to flow over and around the potash coils and germinators. Re- 
duced atmosphere experiments at the same temperature were run 
at the same time, but the correction for water vapor in these series 
makes a direct comparison with the results in hydrogen impossible. 
However, comparison of these hydrogen results with other low 
pressure series run in the same way, makes it possible to draw 
trustworthy conclusions. 
Attempts were made to test the after-ripening of Xanthium 
seeds of different ages, from green to a year old, at normal and 
reduced pressures. These experiments were not extensive, and. 
on the whole not very satisfactory; but the results indicate that 
only very slight changes occur. : 
In all cases control cultures were employed. In measuring 
the growth of the controls, a difficulty presented itself. At full 
atmospheric pressure the roots of Xanthium seedlings penetrate 
the substratum of cotton with innumerable branches, impossible 
of disentanglement and accurate length measurement. In all these 
cases only the unbranched portion of the plant was measured. In 
all the experimental plants the total growth in length was easily 
measured and is so recorded. 
Experiments : 
As indicated above, the experiments proceeded along four lines: 
to determine the minimum oxygen pressure necessary for germina- 
