April 15, 1929
INDUSTRIAL AND ENGINEERING CHEMISTRY
ing the extractives raises the fiber-saturation point. All of these values for the fiber-saturation point are higher than those given by Myer and R e e ~ . Part ~ of the difference between these values and those of Myer and Rees may be due to the difference in the methods used for locating the point on the graph that represents the fiber-saturation point. Myer and Rees determined the points of maximum curvature of their graphs, whereas the present author determined in each case
_97 -
the point a t which the straight-line portion of the logarithmic curve joins the curved portion-that is, the point where the relationship between moisture content and electrical conductivity commences to deviate from the relationship holding for water imbibed by the cell walls ("bound" water). Further differences between the sets of data may very well be due t o the more nearly complete elimination of moisture gradients in this research than in that of Myer and Rees.
A Heavy-Duty Thermostat' D. F. Othmer' UNIVERSITY
OF
MICHIGAN. AAN ARBOR,MIcn
SIMPLE, rugged thermostat was desired t o control at least 4000 watts of electrical energy and maintaiii the temperature of a small boiler constant t o 0.02" C. for days a t a time. The regulating force was the vapor pressure of the water boiling in the apparatus, but any of the more usual types of thermal expansion bulbs might be utilized as well as the bulb of a vapor-pressure thermometer. When the vapor pressure of the liquid may be used, it is preferable because thermal equilibrium is very quickly obtained. Most heavy-duty electrical thermostatic circuits depend on two sets of heaters, one to deliver the largest part of the heat without change and the other to fluctuate with the relay. A single heating circuit is employed in this unit, and because of the design of the relay only one set of leads is necessary for both heating and operation of the relay. The circuit is novel chiefly in its application of the doubleslider Forsythe water-cooled rheostat (supplied by the Central Scientific Company of Chicago) and the single circuit, 220-voltJ a. c. relay (supplied by Eberhach and Sons, of Ann Arbor, Mich.). Two 2000-watt General Electric: immersion heaters were used in the original installation, although subsequently, when the unit was applied to glass flasks, coils of bare nichrome wire were used. Either type is satisfactory because of the very small thermal lag. To the pressure side of the manometer was attached a packing box through which passed an iron rod, nickel-plated and platinum-tipped. A suitable trap prevented liquid from entering the manometer. An ammeter to indicate the current supplied was inserted in the circuit as shown. I n addition to the main switch, an auxiliary switch served to cut out the relay tvhen desired. On examining the diagram it will be apparent that the resistance on the rheostat between the two sliders is always short-circuited by the heavy copper rod on which t,hey operate. When the relay is closed, that part of the resistance to the left of the left slider is also short-circuited, and the only part of the resistance that is still effective is that to the right of the right slider. By varying the positions of the sliders it is possible to obtain any desired value of the current flowhg between the given limit for the open position and any higher value (below the upper limit) for the closed position. In practice, when a change was being made in the amount of heat supplied and the desired current intensity was not known, the two sliders were placed close together near the right of the rheostat. On the open position of the relay practically the whole rheostat was in circuit and only ahout 5 amperes flowed, while on the cloeed position the rheostat was almost entirely cut out and about 20 amperes flowed. The resistance of the heaters was 12.8 ohms, and the varia-
A
1 Presented as part of the paper "The Condensation ot $team" before the Division of Industrial and Engineering Chemistry at the 76th Meeting of the American Chemical Society, Swampscott, M a s s , September 10 to 14, 1928. 2 Present address, Eastman Kodak Company, Rochester, N.. Y.
tion of the heat supplied was between 300 and 5000 watts. This large fluctuation was unnecessary, and from observation of the relative time between makes and breaks of the relay, the rheostat could be adjusted so that the variation was ohly 2 to 4 amperes. For smoothness of operation this was desirable, because there was a larger lag and fluctuation of temperature when the difference between the current flowing with relay open and closed was large than when it was small. Because of the two heavy contact points in the relay, sparking was not excessive even when the difference in intensity of current was as much as 15 amperes.
I
Diagram of Thermostat Connections
This type of manometer varies with change of atmospheric pressure, but the normal fluctuation of the barometer in a few hours is small compared with the vapor pressure of liquids near their atmospheric boiling point. If a very close regulation is desired over long periods of time, even this factor may be eliminated. The left arm of the manometer is connected by tubing to a bulb of a capacity large in comparison with that of the connections. This bulb (not shown in the figure) is installed in the c?nstant-temperature boiler. A vent valve to the atmosphere is connected in the tubing to release the pressure when desired, and is opened when the thermostat is put in operation. After the desired conditions have been reached, this valve is closed and a constant pressure is maintained on the open side of the manometer, which is thus made substantially free from variations due to changes of atmospheric temperature or pressure. This auxiliary compensator is usually unnecessary, but when attached gives a manometer nTith ft vapor-pressure thermometer on one side and a constant-volume gas thermometer an the other.
