ODORS AND THE SENSE OF SMELL. 687 



saucer covered with a thin layer of water, the water immediately 

 retires, sometimes for several centimetres, before the odorous 

 substance. The laws of the diffusion of liquids may be summa- 

 rized by saying that the rapidity depends on the nature of the 

 substance, increases in proportion to the degree of concentration 

 of the solution, and augments as the temperature rises. Graham's 

 dialyzer is based on the very feeble diffusibility of certain sub- 

 stances, like the gums, and the great diffusibility of certain 

 crystalline substances, like salt. It is simply a vessel, the bottom 

 of which is formed of a leaf of parchment paper, that lets the dif- 

 fusible substances pass into the water around it and holds the 

 others. 



The diffusion of gases and vapors, which is more important in 

 questions of smell, is subject to laws which have been only ap- 

 proximately determined. A glass tube about a metre long is 

 used, divided perpendicularly to its length by a thin metallic par- 

 tition, which can be made to slide between two perforated glasses. 

 A gas is introduced into each of the separated halves of the tube ; 

 the supply-cocks are closed, the partition is lifted out, and the two 

 halves of the tube are put in communication ; a half-hour later 

 the partition is shut, and the gaseous mixture contained in each 

 of the compartments is analyzed. Mr. Loschmidt has in this way 

 found the mathematical rule for the measure of the diffusion of 

 different gases, one within the other. 



The volatility of a liquid is expressed by the weight of that 

 liquid which evaporates per second and per square millimetre at 

 a given temperature. All that is known of it is that this weight 

 is proportioned to the excess of the maximum tension of the va- 

 por at that temperature over the tension which it has in the air ; 

 and this weight varies inversely as the atmospheric pressure ac- 

 cording to a law special for each liquid. Evaporation may, there- 

 fore, give us valuable information concerning the purity of the 

 odor, and spare us, in many cases, the delicate problem of deter- 

 mining the maximum tension which is so important a character- 

 istic of substances. A special apparatus has been devised for the 

 rapid measurement of volatility. 



Tables have been prepared showing the relative volatility of 

 different perfumes, of the substances used for adulterating them, 

 and of the adulterations, by means of which a convenient method 

 is afforded for the detection of frauds. 



The influence of different physical forces on the disengagement 

 of odor has been studied ; and possible relations between the 

 colors of flowers and the intensity of their perfumes have been 

 inquired into. It has been found that white flowers represent the 

 largest number of odoriferous species, and after them come red, 

 yellow, green, and blue. The order corresponds with that of the 



