The essential part of the paper explains a new manometer invented by the writer.
The outline of its construction and its action is as follows:-A thermo-couple of iron and constantan is sealed in a small glass tube (dia 0.5 m. m.) of thin wall, the air in the tube being exhausted to a pressure of few centimeter in mercurycolumn. The middle part of the glass tube, where the junction-point of the thermo-coupleis situated, is wound by a fine metal wire which is to be heated by a constant electriccurrent.
Four or more of such a glass tubes with heating coils and thermo-junctions are sealedin a glass vessel (A), dia=1 inch; length: =8 inches, with a branch pipe. All thethermo-couples and heating coils being connected in series with their respective terminalsoutside of the vessel.
Another glass vessel (B) which containes some glass tubes with a heating coil and athermo-couple as mensioned above, and of the same form as the vessel (A) is sealed afterevacuated to a definite vacuum.
The two sets of the thermo-couples in the two tubes (A) and (B) are connected inopposite direction and both the heating coils are connected in series. If we connect theglass vessel (A) to a vacuum to be measured and heat the heating coil by a constant current, the reading of a milli voltmeter which connected to the terminals of the set of the thermocouples continuously gives a measure of the vacuum in wide range with better sensibility andaccuracy.
In this manometer, small flactuation in heating current or appreciable change of surrounding temperature does not effect the voltmeter reading, because both the thermo-couplesin the two glass vessels (A) and (B) are connected in opposition. This is special merit of themanometer. If the vacuum in the vessel (B) be properly pre-determined, the e. m. f. due tothe thermo-couples becomes zero when the vacuum to he measured reaches a certain value.So that by means of the manometer with a sensitive galvanometer connected to the thermocouple circuit, we can know very conveniently whether the vacuum in the tube (A) reachesa predetermined value. In this case the heating current is nearly in different to the indicationof the galvanometer needle.
View full abstract