[Illustration: FIG. 117.—Aeration Apparatus for Testing Lachrymators.] A very satisfactory method for determining lachrymatory values is shown in Fig. 117. Air is measured at _A_ and bubbled through the lachrymatory substance in _B_. The air and gas are mixed in _D_ and pass into _E_, a gas-tight, glass-walled chamber of about 150 liters capacity. The gas is removed through _Ef_, by suction and the volume of the air-gas mixture measured by the flow meter, _F_. After the apparatus has run a few minutes, and the concentration of the gas has become constant, the subject is instructed to adjust the mask, attached at _H_, and to tell whatever he notices just as soon as he notices it. The operator stands in such a position that he can manipulate the stopcock _H_ without being observed by the subject. After breathing air for a time (_H_ is a two-way cock, connected with the air through _J_, and to the chamber through _Eg_) both to become accustomed to the mask and to eliminate, as far as possible, any “psychological symptoms,” the subject is allowed to breathe the gas mixture for a maximum of three minutes. If the expected symptoms are produced in less than this time, the test is discontinued as soon as they develop. [Illustration: FIG. 118.—Type of Spray Nozzles.] For accurate work, it is necessary to work with a pure sample which is at least fairly volatile. Mixtures cannot be run by this method. In this case it is necessary to volatilize each separately, passing the vapors simultaneously into the mixing chamber _E_. A spray method may also be used with satisfactory results. Types of sprays are shown in Fig. 118. ODORS Because of the great value in detecting low concentrations of gases in the field, it is important to know the smallest amount of a gas that can be detected by odor. In some cases, this test is more delicate than any chemical test yet devised. Odors may be divided into two classes, true odors, and mild irritation. By true odor is meant a definite stimulation of the olfactory nerve, giving rise to a sensation which is more or less characteristic for each substance producing the stimulation. Mild irritation defines the sensation which is confused with the sense of smell by untrained observers, but which is really a gentle stimulation of the sensory nerve endings of the nose. This so-called odor of substances producing this effect is not characteristic. Higher concentrations of these compounds almost invariably cause a definite irritation of the nose. Examples of true odors are the mercaptans, mustard gas, bromoacetone, acrolein, chlorine and ammonia. Substances which cause mild irritation are chloroacetone, methyl dichloroarsine, ethyl iodoacetate and chloropicrin. In making the test for odor, the same apparatus is used as for lachrymators. The time of exposure is shortened to 30 seconds, as the subject always detects the odor at the first or second inhalation.