• 78 



bolts, in practically air-tight contact with its bed, which is cut 

 into the case frame. The bottom of the case is of wood, covered 

 inside with a continuous sheet of tin. 



The pump used for forcing air into the case was made by the 

 local tinner at an expense of twenty-five cents. The cylinder 

 of the pump is four inches long and two inches in diameter. A 

 pump of this size will deliver, approximately, one hundred cubic 

 centimeters of air at each stroke, and, at a rate of eighty strokes 

 per minute, it will change the air in the plant-case every thirty 

 minutes. 



The air passes from the pump into a filter-flask, which serves 

 as a " stuffing-box," from which it passes into the series of wash- 

 bottles at a more uniform rate than it would if the pump were 

 connected directly with the wash-bottles. Considerable acid is 

 swept along by the air, from one wash-bottle to the next. This 

 exposes much more, surface of acid to the air, but necessitates 

 the inclusion of a bottle at the end of the series to catch the acid 

 thus carried over. 



An electric fan, within the case, keeps the air well mixed. A 

 thermograph of convenient size, a recording hygrometer, and a 

 small hygrometer of the " Mitthof's " pattern were also kept in 

 the case. This latter instrument is much more sensitive than 

 the recording hygrometer, and thus indicates any temporary 

 variations in the humidity. 



The air in the case can be reduced from saturation to approxi- 

 mately lo per cent, (hygrograph record) in 12 hours, even when 

 the case is full of living plants. In these experiments the air was 

 continually kept as dry as possible, but the humidity could have 

 been maintained at any desired per cent, within a range limited 

 only by the sensitiveness of the hygrometer used, had the index- 

 arm of the hygrometer been made to open and close mercury 

 switches, operating the circuit of the electric motor which drives 

 the pump. The current of dry air would then have ceased 

 when the desired humidity was reached. 



As stated above, the time of death of the plants grown in the 

 case could be determined quite closely. By holding the humidity 

 very low all of the time, the plants dried out at once, and became 



