Page:Popular Science Monthly Volume 41.djvu/704

686 Numerous classifications of odors have been proposed. It is, of course, impossible to quote any rational classification. The natural way is to group around a type, in successive series, odors which resemble one another. Eugene Rimmer has tried to do this in the accompanying table.

The author observes that it would be hard to arrange in any of these series certain peculiar odors like that of wintergreen, or salicylate of methyl and magnolia. Notwithstanding the uncertainties attending the arrangement, we must apparently depend upon classifications based upon this principle for a guide in the study of odors.

All that we know concerning the propagation of an odor is that it consists in an emission of solid, liquid, or gaseous particles. This emission is allied for these three states of matter to the property called diffusion, which consists in the reciprocal penetration at the end of a certain time of the particles of two or more bodies among one another; and also for solids and liquids to the property called volatility, or the rapidity of evaporation.

But little is known concerning the diffusion of solids. If we heat to a high temperature a porcelain crucible within a crucible of plumbago, the plumbago will penetrate the porcelain to a depth varying according to the duration of the experiment. M. Pellat has shown, by delicate measures of quantities of electricity, that metallic surfaces placed parallel to one another a few tenths of a millimetre apart, reciprocally exchange their outer surfaces, as if they emitted a little of their own substance to each other. When the influence ceases, the surfaces gradually lose their foreign coatings, and return slowly to their primary condition.

The diffusion of liquids is easily observed. It can be witnessed by introducing, with a pipette, into a vessel under water a colored liquid, red wine, for example. The wine, being lighter than water, rises to the surface, and does not color the deeper layers of the water till after one or two days. There is doubtless in the complicated diffusion of liquids a kind of chemical action related to the movements on water of camphor and a considerable number of diffusible substances. If we put a bit of camphor on the surface of water, it at once turns round and moves in every direction. If a drop of oil is let fall on the same surface, the movements will cease immediately. The motion arises from the diffusion of camphor in a liquid form on the surface of water. When, after the surface is saturated, there is no more diffusion, the motions cease. They also cease when two currents are produced by different bodies in opposite directions. That there is a liquid diffusion is proved by the fact that when the camphor is placed on a float of pith, or on the polished surface of mercury, there is no movement. So, if a bit of camphor is put into a large