334 Journal of Agricultural Research voi. v. No. s 



In the multiple-point switch the leads from the diflferent groups of 

 thermoelements terminate in a double row of studs arranged in segments 

 of concentric circles, and the galvanometer leads terminate in two metal 

 rings concentric with the studs (PI. XXXIII, fig. i). Metal strips, passing 

 through a vertical shaft at the center of the circles, complete the circuit 

 from studs to rings, the ends of the strips being bent to touch edgewise. 

 On turning the shaft by means of the handle at the top, the strips are 

 shifted from one pair of studs to another, thus connecting the different 

 systems with the galvanometer. The switch includes studs not only 

 for the thermoelement groups described above, but also for resistance 

 thermometers described beyond, so that the same galvanometer will 

 serve for several systems. 



The galvanometer with which the electromotive forces in the thermo- 

 element circuits are detected is a reflecting instrument of the D'Arsonval 

 type, with a coil resistance of 39 ohms. When critically damped, it has a 

 period of 7 seconds, and a sensitivity such that an electromotive force 

 of approximately 2 microvolts in either circuit will cause a deflection of i 

 mm. on a scale i meter from the mirror of the galvanometer. 



With this galvanometer the number of thermoelements in each circuit 

 is sufficient to cause a fairly large deflection when the temperature of the 

 zinc wall is only slightly different from that of the copper wall. In the 

 bottom section, for example, there are 10 thermoelements, the smallest 

 number in any section, each with four differential couples, and each couple 

 having a thermal electromotive force of close to 40 microvolts per degree 

 of temperature difference between the junction at one end and that at the 

 other. All 40 couples being in series, there would be a total electromotive 

 force of 1,600 microvolts for an average difference of i degree between 

 the temperature of the copper wall and that of the zinc wall in this section, 

 or 16 microvolts for an average difference of o.oi degree. Since an 

 electromotive force of about 2 microvolts will cause a deflection of i 

 mm., a difference of o.oi degree would cause a deflection of at least 

 7 mm. It is easy to read a deflection of less than i mm. ; consequently 

 the effect of a temperature difference as small as 0.00 1 degree between 

 the junctions at opposite ends of the thermoelements in this may be 

 observed. The effect of such a difference in the other sections would be 

 greater, because of the larger number of elements; the 16 in the top 

 would cause a deflection of more than a millimeter; the 27 in the upper 

 zone of the sides about 2 mm. ; and the 42 in the lower zone more than 

 3 mm. 



Controlling the Temperature ok the Zinc Wall of the Chamber 



The temperature of the zinc wall is raised or lowered by heating or cool- 

 ing the air confined in the narrow space adjacent to the outer surface of the 

 zinc, which has a corresponding effect on the wall. Parallel with the 

 wall, and about 4 cm. outside of it, is a wall of cork board 38 mm. thick, 



