MEASUREMENT AND CONTROL OF HUMIDITY 



being attempted this system is useless. For large-scale meteorology, machines 

 with built-in controlled cooling, and photocells which detect the moisture 

 film have been used with appreciable accuracy. 



Thirdly, the 'wet bulb' temperature can be used — the temperature to 

 which water is cooled by evaporation into the air. This is altogether both 

 inaccurate and difficult in use. 



Finally, there are methods in which materials are selected whose particular 

 measurable property changes with water absorption. This may be the 

 contraction and expansion of paper, hair, catgut, etc., the degree of ionization 

 of a salt, usually as the conductivity of the electrolyte, the resistance of 

 various porous materials in different states of hydration, etc. These are 

 generally speaking the best methods of measuring humidity. They all suffer 

 from different disadvantages: in particular that they are slow to come into 

 equilibrium, and even more important that they cannot therefore follow 

 rapid changes in humidity, that they have very pronounced hysteresis, and 

 that their calibration changes with age and conditioning; that they are 

 excessively susceptible to contamination of all kinds — particularly to minute 

 traces of grease. They all to a certain extent will lose, or take up, water in 

 the process of coming into equiUbrium, though if the element is made 

 small enough this can be reduced to small proportions, and once equihbrium 

 has been obtained they will at least indicate (a) that a particular humidity is 

 remaining constant and (b) that it has departed from that state to a certain 

 extent. Many small portable measuring instruments such as paper and hair 

 hygrometers have been marketed, in which the movement of the element is 

 directly shown by a graduated scale and pointer, and have limited application 

 over a particular range. This brief account will of course be limited to 

 electrical applications to such instruments. 



CALIBRATION 



The calibration of hygrostats presents an extremely difficult problem and since 

 the only arbitrary standards which can be reasonably achieved are 'dry' air 

 and 'saturated' air the fundamental practice is to produce reservoirs of these 

 (e.g. by drying through a liquid air trap and by bubbling through a sinter- 

 glass filter in distilled water) and to mix in suitable proportions to give 

 the required humidity in a vessel which does not adsorb v/ater vapour — such as 

 polystyrene, p.t.f.e., or silicone-coated glass. 



As a reasonable laboratory alternative, tables exist of the relative humidities 

 of air in equilibrium with saturated solutions of a large number of salts at 

 selected temperatures. So long as care is taken to ensure accurate temperature 

 control and purity of solutions this method will probably provide cali- 

 bration humidities as accurately as any biologist may need. 



PRACTICAL DETAILS OF ELECTRICAL METHODS 



Thermocouples 



The two junctions of a thermocouple circuit can conveniently be used 

 as the wet and dry bulbs of a hygrometer; since they can be made with 

 minute thermal capacity, the wet bulb evaporation can be very small, and 



414 



