A more sensitive modification of an ultra-violet absorption method for detecting mercury vapour is described, cdpable of detecting IO-" g. of mercury vapour in approxlmately 50 ml of air S o sorption of mercury vapour could be detected on l'yrex glass vmsels. Sorption on wheat is influcnced much more by changes in the moisture content than by changes of temperature within the normal range of both faL'cors. Most of the mercury sorbed is LhemiCdlly combined, but with drier grain some can be recovered by airing The reaction of mercury vapour with wheat continues even after two years' expowre to the fumigant. Concentrations of mercury lethal to normal grain weevil eggs are likely to extend for two feet around foci of diffusion in a mass of grain Health hasards appcar t o be slightIn an ear1ic.r paper7 the development of an apparatus for detecting mcrcury vapour in air was outlined, as was the use of the apparatus for experiments on the sorption of mercury vapour on wheat.One serious disadvantage of this detector was its inability to work satisfactorily at and below the range 0.1 mg. of mercury per cubic metre of air. This concentration has been reported as the maximum permissible contamination in which it is considered safe to work for long periods2 Of the possible methods of increasing the sensitivity of the apparatus, that finally adopted was a newly designed amplifying circuit of greatly increased sensitivity.
ExperimentalApparatus D e s i p of amplif3ling circuit.-The circuit was designed to be of high sensitivity and arranged to balance at full light intensity, i.e., with a clean light absorption tube. Any small decrease in intensity at the photocell, due to traces of mercury, caused the circuit to go out of balance very rapidly. As can be seen from Fig. I it was a balanced bridge, and two matched valves were used, opcmting on the linear portion of their characteristics. Particular attention was paid to anodecathodr insulation, and for this reason the QVA $3 photocell was supplied unbased. Thc high resistor, K,, was mounted on porcelain insulators and the metal shielding on the valves rcmovcd to reduce grid-earth leakagch. In operation, any change in the illumination of the photocell, I), resulted in a change in the grid current of V, which was recorded on MI. The current gain between the grid current of V, and MI was of the order of 104.To operate the amplifier after an initial warming up period the H.T. (high tension) voltage was adjusted to ZIO volts as registered by the meter M, using the variable resistor Itl2. The cathode currents in the two valves as shown on meters M A and NIB were then eqixalized using Itll.Any change in the illumination of the photocell P then rcsultrd in a change in the grid current of V, which appeared as a voltage which was recorded on the indicator meter MI. This indicator meter was fitted with a six-position range switch giving three linear ranges, position 3 giving highest sensitivity and positions 2 and I lower sensitivity by switching in resistors R, and R, respectively. Positi...