When a naive onlooker attempts to duplicate the trick, the cream sinks. It should. Cool cream is denser than hot coffee,

meaning that it weighs more per unit volume. The dinner-table performer takes this fact into account before doing the trick and stealthily mixes three spoonfuls of sugar into the coffee, thus making the coffee denser than the cream.

Altering the surface tension of a liquid can also lead to interesting effects. Surface tension arises from attraction between the molecules of a liquid and explains in part why some liquids refuse to mix with others. For example, molecules of mercury are so strongly attracted to one another that they literally squeeze out molecules of water or other substances that do not readily mix with mercury.

A particularly fascinating effect that accompanies a change in surface tension can be demonstrated with an iron nail and a small pool of mercury that is immersed in an electrolyte. The mercury is placed in a shallow container with a spherical bottom, such as a large watch glass, and flooded with about 30 milliliters of tap water. To the water is added four drops of sulfuric acid and about three milligrams of potassium dichromate. (A dozen particles of potassium dichromate about the size of sugar granules approximate this weight.)

A clean iron nail, free of rust, is placed in the solution with the point directed toward the center of the container and the head of the nail overhanging the edge [see Figure 6 ]. Push the nail slowly toward the center of the watch glass until the point makes contact with the mercury. The mercury will immediately dart away from the nail, return, make contact and repeat the performance again and again. Depending on the amount of mercury and the position of the nail, the oscillation will assume an ever changing pattern of amoeba-like forms.

If the pool contains 40 grams of mercury. the nail can be shifted by trial and error into a position at which the mercury oscillates as a standing wave with three crests and three valleys, as illustrated. The action will continue for half an hour or longer, until most of the potassium dichromate combines chemically with the mercury. The apparatus functions as a voltaic cell in which the mercury and the iron serve as electrodes. Current generated by the cell periodically alters the intermolecular attraction of the mercury that accounts for its surface tension. The variations, which are in effect pulses of force, induce the vibration.

The behavior of liquids that float on or within other liquids without much mixing has in recent years attracted the interest of Aristid V. Grosse, president of the Research Institute of Temple University. Grosse and his associates specialize in research at high temperatures. In the course of this work the need arose for a container that could be heated above the melting point of any known substance. Ultimately they devised one: a crucible made of liquid! The furnace that heats the crucible consists essentially of a horizontal, rotating cylinder of metallic oxide exposed on the inside to a plasma jet that develops a temperature close to 30,000 degrees Fahrenheit. The inner surface of the oxide melts; the resulting fluid is held against the rotating shell of solid oxide by centrifugal force. The liquid metal of interest floats on the molten oxide without mixing. (Carbides, nitrides and other metallic compounds can also be used.)

"In the course of these experiments," Grosse writes, "it became apparent that many facts about immiscible or partly miscible liquids had never been investigated. Since experimentation is much easier at room temperature than at high temperature, I decided to study the behavior of a liquid lens of gallium floating on mercury-the simplest case of two immiscible metals that can exist in the liquid phase at room temperature. Gallium is a rare and costly metal, however, a fact that seriously restricted the scope of the experiments.

"For this reason it became apparent that the investigation could be extended only by substituting selected organic and inorganic liquids for the metals. Experiment soon demonstrated that such substances worked nicely-so nicely that I found myself tinkering with them at home. Thus was born my new hobby: making liquid structures. The structures include double lenses, triple lenses and poly-lenses that float on and in water; 'weightless' lenses and other forms that I suspend between solutions, and multiple-layered pillars that stand in liquid. It is also possible to make such structures as enclosures and walls.

"The variety of structures that can be made appears to be limitless. Some combinations behave in strange and unexpected ways. The forces that bind the liquids together still await thorough investigation. Here, then, is a hobby full of opportunity, one I am delighted to share with amateurs.

"In making the structures I use clear glass dishes of various sizes, including beakers, graduated cylinders and tubes. To prevent evaporation all the containers can be closed, preferably by ground-glass covers or ground-glass stoppers. Lipless beakers and crystallizing dishes with covers are ideal. The various liquids are added to the containers by means of pipettes or, in some cases, by disposable hypodermic syringes with a capacity of two to 10 cubic centimeters.

"For preparing solutions I prefer freshly distilled water. Tap water usually contains air that in time comes out of solution and is deposited on the walls of the container as small bubbles. The bubbles obscure the contents of the container. Moreover, when they break free and rise, they set up currents that may disturb the structures. In addition, tap water may contain dissolved substances that react chemically with the liquids. When tap water is used, it should first be boiled and cooled to drive out dissolved gases.

"Some of the chemicals I have used are listed in the accompanying table [Figure 10], together with their density (at room temperature), surface tension and normal boiling point. In general these materials are available from distributors of chemicals, such as the Fisher Scientific Company, Springfield, N.J. 07081. Exceptions include n-cetane, which can be obtained from the Phillips Petroleum Company; silicone oil, from the General Electric Company, Silicone Products Department; fluorocarbon wetting agent FC-128 and other fluorocarbon preparations, from the 3M Company; Freon materials and oil-soluble dyes, from E. I. du Pont de Nemours and Company. The materials are listed in order of increasing density. Those at the top of the list will float on those below and, in general, will not mix appreciably with near neighbors.

"Experiments need not be confined to the listed substances. Many readily available liquids can be used. The density of ordinary paraffin oil, for example, is approximately .9. Hence it floats on water, which has a density of about 1 at room temperature.

"Paraffin oil mixes with carbon tetrachloride, which has a density of approximately 1.6. A solution of paraffin with a density equal to that of water can be made by adding six parts (by volume) of paraffin oil to one part of carbon tetrachloride. The solution can be made slightly denser than water by increasing the proportion of carbon tetrachloride.

"Similarly, the density of water can be increased by making a solution of table salt. Moreover, a solution that increases gradually in density with depth can be made by partially filling a container with warm water and slowly adding cold brine at the bottom through a slender tube that touches the bottom of the container. The brine will diffuse upward, although a uniform concentration would not be attained for years if the solution were undisturbed. In a brine solution drops of immiscible fluids that are slightly denser than water-for example the mixture of mineral oil and carbon tetrachloride-will descend to the depth that approximates their own density. There they will come to rest as free-floating bodies.

"The case of a single lens on water, such as a floating drop of oil, has been intensively investigated, but I have found no reference in the literature to the study of double lenses or poly-lenses. They are systems of lenses denser than water that are supported by adhesion to one or more buoyant lenses at or near the surface. A simple pair is easy to construct by the following steps.

"Add a lens of paraffin oil to a container of water by placing the tip of an oil-filled pipette in contact with the surface near the center of the container and letting the oil flow into a patch about four centimeters in diameter. Add to the water three or four drops of wetting agent (Kodak Photo-Flow 200). With a clean pipette apply o-toluidine to the center of the paraffin oil. The o-toluidine will migrate through the oil lens and form a suspended meniscus that adheres to the lower surface of the paraffin oil. The optical properties of the combination (acting as a lens) can be judged by the accompanying photograph [Figure 5].

"A triple lens can be formed by substituting two milliliters of silicone oil for paraffin oil. To the solution add three drops of a 2 percent (by weight) aqueous solution of FC-128 wetting agent. One milliliter of o-toluidine is added to the bottom of the silicone. The third lens is then made by allowing a few drops of Freon E-5 to migrate through the assembly and form a crescent on the bottom surface of the o-toluidine. If desired, the Freon can be colored blue by mixing in a small quantity of fluorocarbon dye, Type L-1802, which is obtainable from the 3M Company.

"A quadruple lens is built up by a similar procedure. First a double lens is made with two milliliters of paraffin oil and 1.5 milliliters of silicone oil, after adding 10 drops of FC-128 wetting agent to the water. The third lens consists of .8 milliliter of o-toluidine. I stained this lens pink so that its boundaries could be examined easily. The fourth lens was made of .15 milliliter of Freon E-5 dyed blue [below].

"The quadruple lens is a perfectly stable structure. It has eight interfaces: air-oil, water-oil, oil-silicone, water-silicone, water-o-toluidine, silicone-o-toluidine, Freon-o-toluidine and Freon-water. It also has four triple-interface lines: air-water-oil, water-oil-silicone, watersilicone-o-toluidine and water-o-toluidine-Freon.

"The first quadruple lens I made did not last long. The o-toluidine dissolved slowly in the water and disappeared after a few days. I learned to prevent this by first saturating the water with o-toluidine. In some experiments it may be desirable to saturate the water with all the substances used for making lenses. Each lens material can also be saturated with the compounds that border it. To saturate a liquid merely add to it an equal amount of the saturating liquid, shake the mixture, let it stand for a few hours and pour off the supernatant.

"The quadruple lens just described is by no means the highest possible number of lenses that can form a stable structure. The problem of the maximum possible number of lenses is closely related to the problem of the maximum number of insoluble liquids that can be poured one on top of another. The problem has fascinated chemists since the Middle Ages.

"Joel H. Hildebrand, an American pioneer in the study of liquids, first demonstrated that as many as seven immiscible liquids can be deposited as layers in a cylindrical container. In the order of increasing density they are heptane, aniline, water, perfluorokerosene, phosphorus, gallium and mercury. If these layers are completely mixed by being shaken, they will re-form.

"It is also possible to make a system of nine layers in which all adjacent layers are immiscible even though some layers remote from one another are miscible. The substances are, in order of increasing density, paraffin oil, silicone oil, water, carbon disulfide, n-perfluoroheptane, a 62.5 percent solution (by weight) of zinc chloride, tri-n-perfluorobutylamine, liquid gallium and mercury. If a container holding such a system is shaken, the zinc chloride solution mixes with the water and the two fluorocarbon layers combine. The combinations reduce the system to the same number of layers Hildebrand devised.

"The addition of liquid phosphorus from Hildebrand's system would not increase the number of layers because the phosphorus can be mixed with carbon disulfide. This is not to say that Hildebrand's system of immiscible layers cannot be increased. I have devised a system of 13 layers that have been tested in two sections but have not yet been combined into a single column. The liquids, again in order of increasing density, are paraffin oil, silicone oil, o-toluidine, a 2 percent solution (by weight) of common salt, carbon disulfide, n-perfluoroheptane, a 64.5 percent solution (by weight) of zinc chloride, liquid phosphorus, tri-n-perfluorobutylamine, 100 percent phosphoric acid, lead tetramethyl, liquid gallium and mercury This list is by no means the ultimate. Compounds could be synthesized that, when sandwiched between two adjoining layers, would not be miscible with them. For example, liquid heavy-metal perfluoroalkyls would be denser than the corresponding metal alkyls and would not be miscible with them or with liquid gallium. Hence they would fit between layers 11 and 12 on the list.

"It is obvious that in addition to suspended poly-lenses it should be possible to stack lenses upward from the bottom of a container. I call such structures 'reverse' lenses. A simple reverse double lens can be made by first placing 200 milliliters of water containing 2 percent of a 2 percent solution (by weight) of FC-123 wetting agent in a circular crystallization dish 9.7 centimeters in diameter, with a flat bottom and straight sides. A ring of thin aluminum wire is then centered on the bottom of the dish to serve as an anchor that prevents the lens from drifting. By means of a pipette aniline is deposited in the center of the aluminum ring. It forms a beautiful reverse lens 1.25 centimeters in diameter.

"A second lens, of silicone oil, is deposited on top of the aniline. The density of silicone oil is less than that of water. For this reason it tends to lift the aniline lens and to decrease its radius. Aniline has a solubility in water at 20 degrees centigrade of about 3.4 grams per 100 milliliters of solution. As the aniline dissolves, the lens shrinks in diameter. After 24 hours the aniline lens matches the diameter of the silicone lens.

"The area of contact continues to shrink, and at a critical moment the silicone lens pulls the aniline lens upward and rises with it to the surface. One can then see the saturated aniline solution stream downward as a diffraction pattern that resembles in reverse action the 'heat' that rises from a steam radiator. In a few hours the aniline portion of the lens vanishes.

"Having made a number of poly-lenses and their reverse counterparts, it occurred to me that it should be possible to combine them to form a layered pillar that would be self-supporting. To devise a structure of this type I first made a reverse double lens with one milliliter of nitrobenzene and .4 milliliter of silicone oil in 200 milliliters of water containing FC-128 wetting agent, as in the experiment just described. The container was a crystallizing dish 9.7 centimeters in diameter. A double lens was added at the top with one milliliter of the same silicone oil and .4 milliliter of aniline, added by the drop inside the oil with a hypodermic syringe. Two double lenses were then facing each other [Figure 7].

"A thin glass rod was used to push the upper lenses until they were centered directly above the reverse lenses. Finally, the upper lenses were lowered into contact with the reverse lenses by slowly draining water from the container through a siphon fitted with a pinchcock. The resulting structure, which was held together entirely by the forces of surface tension, measured about 2.5 centimeters in height. In the course of a few hours the aniline lens decreased in size because of its solubility.

"The force per unit area that tended to pull the structure apart finally exceeded the effective surface tension and the pillar separated. Then the aniline dissolved completely, leaving only the upper silicone lens. The reverse lens remained intact because the solubility of nitrobenzene in water is only .19 gram per milliliter in water at 20 degrees C.

"The tallest liquid pillar I have made so far consists of five layers with a total height of eight centimeters. The structure was self-supporting. It was assembled in a beaker of two-liter capacity and required about two days of patient work.

"The bottom layer, which consists of carbon disulfide, was deposited in a thin ring of glass to prevent drifting and was surrounded by an 18 percent (by weight) solution of table salt to increase the buoyancy of the carbon disulfide. At the time of the experiment I had no access to a silicone oil of a density intermediate between that of o-toluidine (.999) and nitromethane (1.130). Hence I increased the density of the then available silicone oil by adding carbon tetrachloride in the amount of 17 percent (by volume).

"The five-layered pillar lasted many hours. It was, however, inherently unstable because the carbon tetrachloride in the middle layer dissolved slowly in both adjacent layers, with an unexpected effect on surface tension at its boundary. A small 'tongue' of o-toluidine periodically moved down from the center into the silicone, in what can be most clearly described as a 'licking' motion. The amplitude of the motion increased in the course of a few hours. Eventually a cylinder of o-toluidine formed completely around the silicone, which then assumed the form of a beautiful hyperboloid that lasted for about an hour. In time the hyperboloid separated at the waist. The silicone drifted upward, breaking the pillar, and joined the other silicone lens at the surface.

"Multiple-liquid lenses can also be formed at the plane boundary between two liquids. One of many possible examples can be prepared by first pouring cetane and perfluoroheptane into a container five centimeters deep. These liquids separate into two clear, immiscible layers each two centimeters thick. At the boundary insert by pipette one milliliter of o-toluidine. The liquid promptly spreads between the layers to form a classical lens.

"With a hypodermic syringe inject by the drop, at the lower boundary of the o-toluidine, .5 milliliter of water stained deep red with a water-soluble dye. A dozen or more independent droplets should be thus attached to the bottom of the o-toluidine lens. After some hours the drops will coalesce to form the second lens.

"The double lens is merely a special case of what I call 'weightless' structures: systems of liquids suspended between layers of liquids that differ in density, or within a single layer of variable density as represented by the saline solution previously described. To make a colorful triple lens of this type, insert 400 milliliters of cold brine solution below 600 milliliters of warm distilled water in a tall container, such as a one-liter beaker. Put one or more large drops of o-toluidine, dyed pink, in the water. The drops tend to float to the surface. To each drop add with a hypodermic syringe a few small drops of Du Pont E-4 Freon stained blue with Huorocarbon dye. The slightly heavier lens so formed will sink into the zone of diffusion between the salt water and the fresh water.

"The triple weightless lens is now completed by injecting (with a clean hypodermic syringe) colorless silicone oil into the o-toluidine. The buoyancy of the oil causes the third lens to form as a cap on top of the assembly. The triple lens will seek that depth in the zone of diffusion at which the density of the brine equals the average density of the components in the triple lens. By adding silicone or Freon as desired the lens can be adjusted so that it will come to rest at any given depth in the zone of diffusion and will remain there for hours.

"The thickness of the zone of diffusion can be increased by stirring the brine gently and letting the solution come to rest. Spheres in a range of sizes, densities and colors, when injected into the solution, come to rest at various depths and can serve as attractive, although only roughly approximate, models of planets and other astronomical objects in space. It is interesting to speculate on how structures of this type would behave under the weightless conditions of outer space.

"What good are poly-lenses? Aside from their value as entertainment and the beautiful opportunity they provide for investigating surface phenomena, it turns out that single liquid lenses are excellent devices for concentrating solvent-insoluble impurities. For example, an o-toluidine lens with a volume of two cubic centimeters was allowed to shrink in water at room temperature to a sphere one millimeter in diameter-a volume of .0005 cubic centimeter. The concentration of the impurities was thus increased by a factor of 4,000!

"Some of the materials used in these experiments are toxic and must be handled accordingly. Work in a well-ventilated room. Wash your hands frequently with soap in running water and at all times keep them away from your mouth and eyes."

Bibliography

THE SOLUBILITY OF NONELECTROLYTES Joel H. Hildebrand and Robert L. Scott. Reinhold Publishing Corporation, 1950.

DOUBLE, TRIPLE AND POLY LENSES OF ORGANIC LIQUIDS ON AND UNDER WATER. A. V. Grosse. Research Institute of Temple University, 1967.

The Society for Amateur Scientists (SAS) is a nonprofit research and educational organization dedicated to helping people enrich their lives by following their passion to take part in scientific adventures of all kinds.