JUST as the accurate measurements of controlled phenomena have taught us much that is definite and much that is useful about the divisibility of matter, so also cleverly designed and skillfully executed experiments are giving us equally definite knowledge of the atomic nature of electricity, of the fact that there is a measurable and seemingly ultimate limit to the divisibility of an electric charge, and what the exact magnitude of the ultimate charge is. Like nearly every other discovery of importance, this one, too, has a long and an honorable history. In fact it can be traced back, step by step, almost a century. The first, and in many ways most fruitful contribution to this subject, was made by that prince of experimenters, Michael Faraday. It consisted in proving that when an electric current is passed through a water solution of any one of certain substances, the substance itself is taken out of the solution, to an extent dependent entirely upon the quantity of electricity so passed, and upon the nature of the substance dissolved. Thus, when the same current is passed in series through the water solutions of several salts, such as sodium chloride, silver nitrate, copper sulphate, and the like, the weight of the metal deposited, or, under proper conditions, permanently removed from the solution, is directly proportional to the atomic weight of the metal itself, and inversely proportional to its valency, that is to say, to the number of hydrogen atoms necessary chemically to take the place of one atom of the metal. This proportionality between the quantity of el'2ctricity passed through the solution, and the resulting amount of chemical decomposition, holds rigidly true, within the limits of experimental error, under all conditions, and hence it seems practically certain that to each ion in a solution conveying an electric current there belongs a definite electrical charge; and that the smallest of such charges is that carried by a single hydrogen atom, or by a single atom of any other univalent substance. A bivalent atom, such as copper in copper sulphate, carries just twice the smallest, or univalent charge; a trivalent atom three times the smallest charge; and so on for atoms of still higher valency. Hence, in electrolytic solutions there is a measurable smallest possible charge of which larg ?r electrolytic chargps are only definite multiples. A natural inference from these experimental facts would be (and this inference was drawn by many) that electrical charges are just as definitely atomic in their nature as, for instance, is a mass of iron. That just as, under given conditions, there is a limit to the actual, though not to the conceivable divisibility of matter, so too there is a limit to the actual, though again nnt [c the conceivable divisibility of an eledrh charge. This, however, was only an inference, and for many years the way to test it, in the case of any quantity of electricity o:.her than that used in the decomposition of an electrolyte, was not evident. Besides, even in the process of electrolysis, the most refined measurements could directly detect nothing less than the aggregate of countless millions of elementary charges, so that the value of the unit charge was only inferen-tially and not immediately determinable. About a dozen years ago, J. J. Thomson, H. A. Wilson, and C. T. R. Wilson began, in the Cavendish laboratory at Cambridge, England, a series of most ingeniously devised and skilfully executed experiments that proved the existence of minute electrical charges in conducting gases, and showed their average value to be, as nearly as could be determined, the same as that of the electrolytic charge spoken of above. C. T. R. Wilson showed that a fog of water particles will form in dust-free air whenever the degree of super-saturation is sufficiently pronounced. If the air is ionized, or has been rendered conducting through the action of X-rays, or otherwise, then a four-fold Buper-saturation causes condensation of the water vapor on the negative electrons; a six-fold super-sat- uration gives condensation on the positive electrons; and an eight-fold on the neutral molecules of the gas itself. Now, a knowledge of the amount of water vapor present, and of the extent of the cooling below the dew point by which the super-saturation is produced The apparatus by means of which Prof. Millikan has isolated an ion and measured its charge. enab1es one easily to compute the weight of the water condensed as fog. Furthermore, if the fog is left to itself, it slowly settles at a rate which, as Stokes proved long ago, depends upon the size of the individual particles and upon the viscosity of the medium through which they fall. A measurement, then, of the rate at which the fog falls, since it all falls at about the same rate, enab:es one to calculate the size of the individual particle, and this knowledge of the siz- of the particle, together with a knowledge of the amount of water condensed, at once gives the total number of particles. On bringing a charged body near this electrified fog its motion is modified. and a means is at hand for meas-uring the magnitude of the charge on each partic'e. PrObably the simplest method of measuring this charge, through the behavior of the fog as a whole, is that devised by H. A. Wilson. The rate of fall of the fog is measured when there is no external electric feld act'ng on it, and thus the size and weight of the individual particle determined. After this, a vertically [ljrected uniform electric field is brought to bear on the particles, and regulated to just counteract the force of gravity, so that the fog neither riSES nor falls. Under these conditions of equilibrium the value of the charge on each particle of fog, multiplied by the s'rength of the field, is equal to the weight of the suspended particle, and hence ,vhen both the weight of the droplet and the strength of the field that keeps it in suspension are known, the numerical value of tlle cllarg3 is also known. All this, however, assumes that the rate of fa'l of the fog en ma ss e, the group rate, is th8 same as would be that of a single or.e of its droplets if alone. This, as a 1 tter of fact, is not rigidly true. For this, and for other 1eaSOIs too, 1: se3med extremely desirable to Prof.' R. A. Mi1l!l'an, of the Ryerson physical lOlbo:atory, at the University of Chieago, somehow to iso:ate and to measure an ion entirely by itself. A f<w years ago it would have been almost silly even to have dreamed cf accomplishing such an experimental feat. But Prof. Millikan !has most cleverly solved this seemingly impossible problem, and by so doing probably has made by far the most accurate of all determinations of the value of this fundamental unit, the atom of electricity. Xor is this all, for the improved value of this unit carries with it correspondingly corrected values of other things -such as the number of atoms per unit weight of any given element, and the mass of the individual atom itself. The full aceount of this ingenious an(l valuable investiga t ion is given in the Physical Review for April of this year, and should be consulted by any one especially interested in the subject. In brief the process by which single ions were isolated and individually measured was as follows': A fine spray of oil was blown by dust-free air into a dust-free chamb', the bottom of which was closed by a brass disk 22 centimeters in diameter. The center of this disk, which was perfectly plane on the under side, was pierced by a pin-hole, through which an occasional oil droplet fell. Strictly parallel to this disk, and just 16 millimeters below it, was another brass disk of the same size, A band of thin ebonite was bound around the edges of the disks, while ebonite rods kept them fixed in position and also strictly insulated from each other. In this way a cylindrical air chamber, 22 centimeters in diameter and 16 millimeters long, was formed between the two parallel brass plates. A parallel beam of light was passed through two diametrically opposite glass covered holes in the ebonite band, and hence immediately beneath the pin-hole in the upper plate. Through a third glass covered hole in the ebonite band a low power telescope was so focused as to show distinctly any object floating in the air immediately beneath the pin-hole. As soon as one or two dro,Jlets chanced to fall through this opening, and, therefore, into the field of the telescope, it was closed by an electromagnetically operated eover, so as to prevent, as far as possible, all disturbances due to air currents. Changes in the size of the drop were almost entirely eliminated by the use of substances that evaporate slowly, and by the additional precaution of having the volume of the cylinder, through moistening its walls, already saturated with the vapor of the substance used. The rate of fall of the droplet, due to the force of gravity alone, was measured, and in this way its size and mass ap'Jroximately determined, as above explained. The plates were then charged to a known difference of -lectrical potential and hence the move· 206 SCIENTIFIC AMERICAN September 2, 19 J I ment of the droplet, if electrified, was changed, The new velocity was measured and its direction noted, These measurements with the electric field alternately off and on were repeatedly taken, but during the course of the observations it frequently happened that the droplet encountered and entrapped a free ion of one or tpe other sign, as was evidenced through its abrupt changes in velocity, The more ionized the gas, the more frequent the captures, By thilS process free gaseous ions of either sign have been captured at will, either singly or in multiples, and their magnitude has been so carefully measured, under conditions so free from assumptions, that the size of the electrical atom, the smallest quantity of electricity now attainab1e, is known probably to within one part in 500 of its actual value. Numerically this value is the absolute electrostatic unit multiplied by 4.891 X 10-1°, a quantity inc om· prehensively small. To iliustrate its exoessive minuteness, let us suppose one hundred millions of people should begin to collect and count these atoms of electricity, just as Prof. Millikan has collected and counted them, only at the extraordinary rate of 100 evel'y minute, and let them keep steadily at 't until they had enough to generate by electrolysis sufficient hydrogen to fill a child's toy baUoon of some eight inches diameter. Certainly this would not be enough elec· tricty to consider in any commercial transaction, and yet it would take the above hundred million hustlers geological ages-in fact something like four millions of years-to coned it by the process described. Surely, then, the electric ion is small beyond comprehension, and its de,fnite isolation and exact measurement stands forth as one of the cleverest, as it is also one of the most important, achievements of modern physics.