A McLeod gauge cannot be used with gases, such as chlorine, which are chemically very active and attack mercury, so various types of membrane manometer have been devised. Very thin spiral glass tubes, working on the principle of the Bourdon gauge used for steam pressures, have been used for such gases, the move ment being measured with mirror and scale : such gauges are not sensitive to pressures less than •I mm. of mercury or so. Scheele and Heuse have used the mechanical deformation of a very thin metal membrane to measure small pressures. One side of the membrane is exposed to the gas or vapour, while on the other side the vessel is evacuated as thoroughly as possible by a high vacuum pump. The very small movements produced are meas ured by an interference method, and results have been obtained down to .000I mm. of mercury, but the instrument is troublesome to work with.
The exceedingly low pressures produced by modern technique have led to the construction of a whole class of new manometers, depending for their action on the anomalous behaviour of gases which begins to manifest itself as soon as the mean free path is comparable with the linear dimensions of the vessel. The con ductivity and the viscosity are properties which are independent of the pressure so long as the mean free path is small, but change rapidly with the pressure when the pressure is very small, becom ing proportional to the pressure below a certain limit of pressure, depending upon the nature of the gas and the dimensions of the vessel. Again, at low pressures, the radiometer effect manifests itself ; that is, there is a mechanical repulsion between a warm surface and a colder surface opposite it when gases at low pressure he between them. (See RADIOMETER.) Further, with an appa ratus of the type used by Foote, Meggers, and Mohler for measur ing resonance potentials (see RESONANCE POTENTIAL), so long as, first, the pressure is low enough for the electrons to have a free path between the hot wire and the surrounding grid, and, secondly, the potential fall accelerating the electrons is well above the ionization voltage, the ionization produced between the grid and the surrounding cylinder is proportional to the pressure. All these effects have been utilised in the construction of special low pressure gauges.
Two types of viscosity gauges may be mentioned. In one, devised by Langmuir and Dushman, a very light horizontal disc of mica hangs suspended by a quartz fibre directly above a second horizontal disc. This second disc is carried on a vertical shaft which can be set in rapid rotation by a rotating magnetic field es tablished outside the glass vessel. The drag of the rotating disc on the suspended disc, which is proportional to the speed of rotation and to the pressure of the gas between the discs, is measured with the help of a mirror attached to the sus pended disc. The effective viscosity depends not only upon the pressure but also upon the nature of the gas, so that the gauge re quires special calibration. On account of secondary effects it is not reliable below a certain low pressure, of the order of 5 X microbars of mercury, and on account of the lack of proportion ality of drag to pressure it cannot be used at pressures above .02
MM. of mercury. It has not found very wide employment.
A much simpler form of viscosity gauge, due to Langmuir, consists merely of a vertical quartz fibre, rigidly fastened at the upper end to the top of the vessel. When the vessel is tapped the fibre is set into vibration, the amplitude of which decreases at a rate determined by the pressure of the residual gas in the vessel. The gauge can be used down to pressures of .05 microbar.
The conductivity gauges, sometimes known as Pirani gauges after the deviser of the first model, or Pirani-Hale gauges since C. F. Hale introduced improvements, depend for the measurement of the conductivity upon the fact that the electrical resistance of a wire varies markedly with the temperature. The gauge is very similar to an electric lamp, consisting of a filament of pure platinum wire, which has a large temperature coefficient of resistance, enclosed in a glass bulb connected to the vacuum. The wire is put in one arm of a Wheatstone bridge, the current flowing through it raising its temperature by ioo° C or so. The exact temperature at which it stands, and so its resistance, will depend upon the conductivity of the gas, which, for a given gas, measures the pressure. The change of resistance consequent upon the intro duction of the gas may be measured, or, better still, as pointed out by Norman Campbell, the voltage change required to keep the filament at constant temperature, since calibration can then be carried out at one temperature.
The reading of a viscosity gauge or of a conductivity gauge depends upon the nature of the gas remaining in the vessel, since the physical properties in question are functions of the atomic weight as well as of the pressure. The magnitude of the radiom eter effect is, however, independent of the nature of the gas, and the gauge constructed by Knudsen upon the radiometer principle is therefore often known as Knudsen's absolute manometer. If of two parallel strips one be maintained at an absolute tempera ture the other at an absolute temperature which is that of the rest of the vessel, then the repulsion between the two plates due to radiometric forces is 1p (v Ti/T2- I) where p is the pres sure of the gas (see RADIOMETER). This repulsion can be meas ured with various experimental dispositions. A moving plate, of mica, glass, or metal, may be carried at one end of a short arm suspended from a torsion wire, with a small counterbalancing weight at the other end of the arm, and either a small part of the wall of the vessel, parallel to the plate, heated externally by a current of air, or by water, or the whole of the vessel heated, and the plate protected by an inner tube from molecules coming from all parts of the wall but that opposite one side of the plate. In another form electrically heated metal strips, introduced inside the vessel, are used as warm plates. The gauges are sensitive to pressures as low as a few times mm. of mercury. The trouble with all the radiometer gauges is that it is difficult to be sure of the temperature of the cold plate, and the gauges all have to be calibrated by the introduction of measured minute quantities of gas into the highest possible vacuum.