STUDIES IN PLANT RESPIRATION AND PHOTOSYNTHESIS. 27 



wire. This copper wire floats on the mercury in the trough. The 

 sides of the trough and the rest of the wood were covered with several 

 coats of Bakelite paint to insure insulation. The trough is parti- 

 tioned off with thin pieces of mica, dividing the whole trough into 

 quadrants which are completely insulated from each other. If 

 necessary the trough can be divided into a larger nuniber of sections. 

 The mercury in each section (in fig. 4 there are four) is connected 

 by means of a short wire to a binding post V. A wire from each one 

 of these binding posts connects to one of the electromagnetic coils, 

 the other end of the coil connecting to the line. Thus the circuit 

 is completed, the current running from the line to binding post 

 L, in turn through the mercury cup on the clock, the arm (not through 

 the clock !) , through a portion of the mercury in the trough, through 

 a sectional binding post V, to a coil, and back to the line. 



As the clock moves, the arm is carried over the trough, making 

 contact in one section of the mercury. When the arm reaches a 

 mica intersection it breaks the connection with one section and makes 

 contact with another. Thus the current is broken from one coil 

 and run through another coil, which process closes one valve and 

 opens another. In this manner a 24-hour clock with four intersec- 

 tions makes possible four 6-hour periods. The periods can be in- 

 creased or shortened at will by changing the number of intersections 

 or by using a clock of different rotation. For this work commercial 

 110-volt A. C. current was used; the coils were constructed with a 

 resistance of about 80 ohms and heated very little. Sparking can 

 be avoided by insertion in the circuit of suitable condensers, although 

 no trouble was experienced from this source. 



From the mercury valves the air-stream passes to the absorbing 

 tubes. These were 10-bulb Meyer tubes which were supported 

 by wooden racks. The glass tubes were connected by means of 

 heavy rubber tubing and all rubber connections were wired and 

 covered with Bakelite paint. Beyond the absorbing tubes a pressure 

 regulator was inserted. This was patterned after the well-known 

 Palladin regulator^ and modified so that a cylinder containing a 

 saturated solution of calcium chloride was used as the liquid. This 

 permitted more accurate adjustment of the pressure than mercury. 

 A water manometer was also inserted through the rubber stopper 

 in the cylinder, so that any change in the pressure in the system 

 could be detected at once. The pressure regulator was immersed 

 in the water of the thermostat to avoid the influence of changes of 

 temperature. With this device the pressure was kept quite con- 

 stant, never exceeding 2 cm. of water, and the air-stream was per- 

 fectly regular. Connection was made from the pressure regulator 



1 Abdebhalden, E. Handbuch der Biochemischen Arbeitsmethoden, vol. iii, 1, 481 (1910). 



