Balance, Heavy Liquids, & Immersion Liquids

Since this assignment is concerned in part with specific gravity determinations by both heavy liquids and the diamond balance, the following general discussion on the theory of S.G is in order.

The quantity of matter in a given space is referred to as Density. Expressed scientifically, it is the mass of any substance contained in unit volume. Thus, the mass, or weight, in a cube of lead is much greater than the contained in a cube of wood of the same size. The densities of gemstones depend on the closeness with which their atoms are packed and on the weight of those atoms. A one-carat zircon is much smaller than an amethyst of the same weight. To say that a zircon is heavier than an amethyst conveys little information; to arrive at an exact comparison, we must compare the weight of equal volumes of each stone with the weight of an equal volume of water. The number of times a stone is heavier than the weight of its own volume of water is called its SPECIFIC GRAVITY (also written S.G. or Sp. Cr.). Thus the specific gravity of a gemstone is the ratio of its density to the density of water.

Specific gravity is technically defined as the ratio obtained by dividing the weight of a body by the weight of an equal volume of distilled water at a temperature of 4° C. (39.2° F.). Water at 4°C has a maximum density and hence can be used as a standard, but in gem testing it is usually possible to obtain sufficiently accurate results without considering temperature.

The specific gravity of quartz is 2.65, which means that a quartz gem of a given size weighs 2 and 65/100 times as much as a volume of water of the same size. Specific gravity is represented by a number that is definite for each species, within certain narrow limits. Hence, its determination offers a sure and ready test in the identification of un-mounted fashioned gems without injuring them.

Specific gravity is an important property, not only for the purpose of identification, but because of its effect on the relative size per carat of gems. For example, a one-carat quartz is considerably larger than a one-carat diamond, which, in turn is larger than a one-carat zircon. Stated another way, the higher the S.C., the smaller the stone per carat, and vice versa. For this reason, when replacing a setting in a ring, a jeweler must give the stone's size in millimeters, rather than in weight, if he is replacing the original stone with one of another species. Many jewelers become adept at estimating the weight of a diamond by appearance; few, however, are equally proficient in estimating the weight of colored stones. To do so requires a general idea of the relative specific gravities of the various stones. As a further example, a one-carat amethyst is much larger in fact, perhaps two-thirds greater in diameter than a zircon of equal weight.


General Discussion.

The use of liquids for the determination of specific gravity was discussed in some detail in the Colored Stone Course. It is a valuable test, and occasionally may even warrant the preparation special mixture of liquids to solve the problem at hand, if the time and expense is justified by the situation.

The following liquids are useful for one purpose or another by the gemologist. Some are available readily, and others can be obtained only from chemical-supply houses or the Gemological Institute. Densities are given only for those liquids normally used for specific- gravity estimations.

Fluid Specific Gravity S.G.
Acetic Acid 1.052
Acetone 0.787
Acetylene, liquid 0.62
Acetylene, liquid 0.38
Adipic acid 0.72
Alcohol, ethyl (ethanol) 0.787
Alcohol, methyl (methanol) 0.791
Alcohol, propyl 0.802
Ammonia (aqua) 0.826
Aniline 1.022
Benzene 0.876
Benzil 1.084
Bromine 3.12
Butane, liquid 0.601
Caproic acid 0.924
Carbolic acid 0.959
Carbon disulfide 1.265
Carbon tetrachloride 1.589
Carene 0.86
Oil, Castor 0.959
Chloride 1.56
Chloroform 1.469
Citric acid 1.665
Coconut Oil 0.927
Cotton Seed Oil 0.929
Cresol 1.027
Creosote 1.07
Crude oil, California 0.918
Crude oil, Mexican 0.976
Crude oil, Texas 0.876
Cumene 0.862
Decane 0.728
Dodecane 0.757
Ethane 0.572
Ether 0.716
Ethylamine 0.683
Ethylene glycol 1.1
Fluorine (freon) refrigerant R-11 1.48
Fluorine refrigerant R-12 1.315
Fluorine refrigerant R-22 1.197
Formaldehyde 0.815
Fuel oil 0.893
Furan 1.421
Furforal 1.159
Gasoline, natural 0.713
Gasoline, Vehicle 0.739
Glycerin 1.263
Glycerol 1.129
Heptane 0.681
Hexane 0.657
Hexanol 0.813
Hexene 0.673
Hydrazine 0.797
Kerosene 0.82
Linolenic Acid 0.902
Linseed Oil 0.932
Mercury 13.633
Methane 0.466
Milk 1.035
Naphtha, Petroleum Naphtha 0.667
Wood 0.701
Napthalene 0.963
Nonanol 0.823
Octane 0.701
Olive Oil 0.703
Oxygen 1.14
Palmitic Acid 0.853
Pentane 0.755
Phenol 1.075
Phosgene 1.381
Phytadiene 0.826
Pinene 0.858
Propane 0.585
Propane 0.495
Propylene 0.516
Propylene glycol 1.036
Pyridine 0.968
Parole 0.969
Resorcinol 1.272
Sabiname 0.814
Sea water 1.028
Silane 0.719
Sorbaldehyde 0.898
Stearic Acid 0.941
Styrene 0.906
Terpinene 0.85
Toluene 0.865
Turpentine 0.871
Water, pure 1
Water, sea 1.025

The density of bromoform and methylene iodide may be reduced by adding xylene, and Cleric's solution by adding water. These mixtures permit the preparation of any desired density up to 4.30.)

Little was said in previous assignments about the maintenance of heavy liquids. Mixtures of different liquids used to obtain densities useful for special purposes, such as a combination of bromoform and xylene to separate emerald from synthetic emerald, may change in density as a result of unequal evaporation of the constituents. Although xylene is somewhat more volatile than bromoform, they are so nearly identical in this respect that mixtures of the two are relatively stable. An uncapped bottle is likely to lose 50% or more of its volume by evaporation before the S.G. increases appreciably. Mixtures of xylene and methylene iodide are much less stable; they must be adjusted (by the addition of a drop of xylene) more often. The higher cost of methylene iodide also makes it less suitable than bromoform for preparing liquids with densities below that of pure bromoform; i.e. 2.85. Because of unequal rates of evaporation, liquids of this type must have indicators of known S.G.'s to show that they are still at the required density at the time they are used. The emerald liquid, for example, which is adjusted initially to a density of
2.67, is functioning if the synthetic emerald barely floats and the other indicator (calcite) sinks slowly.

Practical liquids maintained in the laboratories include dilated bromoform at 2.57, 2.62 and 2,67; bromoform at 2.85; a dilution of pure methylene iodide at 3.05; and pure methylene iodide at 3.32. Dilutions of pure Clerici solution at 3.52 and 4.00 are sometimes used. The 2.57 liquid serves to distinguish microcline and orthoclase feldspar from chalcedony quickly. The 2.62 liquid distinguishes Cryptocrystalline quartz from crystalline quartz. The 2.67 liquid separates most synthetic from natural emerald. The 3.05 is used for several purposes, but it is often called the tourmaline liquid, because
most tourmaline varieties just barely float in it, whereas spodumene, andalusite, and other stones that may resemble tourmaline sink in it. A 3.52 liquid is used for separating diamond substitutes from diamond, and the 4.00 liquid is called the ruby or sapphire liquid. These last two liquids, made from Clerici solution, have little application in testing, since most stones falling in this range of S.G. usually can be tested more easily by other methods. Also, thallium salts, constituents in Clerici solution, are poisonous and highly corrosive; thus, they should be used with reasonable care and are not recommended for general use.

One other type of material is available that has a possible application in specific-gravity determinations. This is a suspension of fine-grained heavy solids in a heavy liquid. Just as a ferrosilcon suspension in water can be used effectively to perform as a heavy liquid of 2.85 density in the gravity separation of diamonds and other dense minerals from blueground, so suspensions in heavy liquids can be prepared that have effective densities anywhere in the range of 3.33 to about 7.0. However, these have the distinct disadvantage of being then, opaque muds; therefore, a stone tested in this manner must be fished for if it sinks. If the stone is small, many minutes may be required to locate it with tweezers. The alternative is to pour
the suspension through a screen, but this means added equipment to clean and the rapid loss of the suspension adhering to the container and screen. The only advantage to suspensions is the high density that is possible.

 The Use of Liquids For Estimations.

The behavior of a stone in a liquid should enable the gemologist to judge specific gravity fairly accurately. Not only is it possible to determine between what two liquids the S.G. of a stone falls, but the rate of sinking in the lighter liquid, as compared to the rate of rise when held and then released from the bottom of the next heavier liquid, should enable a close approximation of the S.C. to be made. If one could estimate accurately the proportion of the volume that remains in the liquid, it would be possible to determine the S.G. of any stone floating in a liquid. For example, if a stone is nine-tenths submerged in a 3.00 liquid, it has an S.G. of nine-tenths of 3.00 or 2.70. Such estimations cannot be made accurately on cut stones, however, due to their irregular shapes.


Certain precautions must be observed when using high-density liquids for S.G. determinations. Be sure that the indicators are in the right position (i.e., floating or sinking, as the case may be); if not, it is necessary to adjust the liquid. For example, if both indicators are at the bottom, the heavier ingredient must be added. If both are floating, the lighter must be added. If it is necessary to add the heavier liquid usually considerably more is needed to effect ) the resetting of the liquid's desired density than if the lighter liquid is being added. The reason for this is that the predominant liquid will almost certainly have the higher S.G. As a consequence, it takes many more drops to change the density appreciably than when the lighter constituent, the S.G. of which is greatly different from that of the heavy liquid, is added. Thus, if you are attempting to reduce the S.G. of the liquid, add a tiny drop at a time and stir it between each drop. Often, one drop of the lighter one will be sufficient. If too much is added, a time-consuming re-correction is necessary. It should be obvious from this that care is necessary to avoid contamination of one liquid by mixture with that from a heavier or lighter one. Be sure that all liquid is wiped from the stone and the tweezers before the stone is placed in another bottle.

There are some circumstances in which the use of some of the immersion liquids is ill-advised. Most of these are obvious, but not all of them. Almost anything suitable for cutting as a gemstone should be safely immiscible in water. Methylene iodide or bromoform, on the other hand, must be handled with greater care. Very porous turquois of low S.G. may absorb a significant amount of any liquid in which it is immersed; this could, cause some discoloration. With naturally colored densest turquois of fine quality, this is not a problem. Even with the porous material, it should be possible to remove the liquid from the pores with a solvent, but it is unwise to use these liquids for highly porous stones.

Discoloration is a possible result also when light colored stones with cleavages or fractures that open to the surface are immersed in liquids with strong color. Again, it can be removed. If it is a necessary link in the chain of tests, the specimen should be dipped in xylene or another solvent of methylene iodide and/or bromoform and dried carefully.

A third source of possible difficulty is a recent resin such as amber, copal, Kauri gum or other substitutes that may be attacked by bromoform, carbon tetrachloride or other solvents to the extent of softening and loss of luster. When block amber is faked by adding insects and filling it with another resin, the added resins often soften in solvents may or may not be diamonds, the tendency may be to check quickly by immersion. If this is done, it is safest to use water.

Sometimes, pale aquamarine beads are coated thickly with a green plastic to simulate emeralds. The natural inclusions in the beryl core are obvious to the eye, so this makes an effective imitation. The plastic may be attacked by bromoform, destroying the luster.

Triplets may suffer from liquid attack on the colored cement layer. It is unlikely to be noticeable at once, but damage may occur.

Although such damage to a stone is very rare, the jeweler gemologist must be exceedingly careful not to damage customers property; therefore, some of the pitfalls have been pointed out.

In general, liquids containing the haloid elements (i.e., chlorine, bromine or iodine) are hard on the skin. Be sure to wash your hands immediately after using the liquids. Clerici's solution is even more dangerous. Any portion of the finger that comes in contact with it turns white, and the white area will be visible for days. Most of the liquids are poisonous, if taken internally. Again, Clerici's solution is a prime offender.

The liquids containing bromine or iodine are sensitive to light, for a chemical breakdown occurs that releases these elements, darkening the liquid to near opacity; therefore, the liquids should be kept in the dark when not in use. In addition, copper filings should be kept in the bottom of the container, to react with what bromine or iodine is released; these are provided in the standard sets of R.I. liquids. If care is taken, the liquids will remain transparent for a prolonged period of time.

When the liquids are used frequently, they tend to become filled with lint. Without too much loss of liquid, it is possible to remove the lint by pouring it through filter paper.


The Various Liquids and Their Functions

Since faceted gemstones are designed to reflect light with maximum efficiency, their brilliancy makes it difficult for the gem-tester to examine them effectively under many conditions. For this reason, it is necessary to surround them with something that reduces materially their ability to reflect light and to make them as easy to see into as a pane of glass. There are many purposes for which this technique is useful.

Today, a synthetic corundum made by the Verneuil process is occasionally totally without visible inclusions under all practical magnifications. When a stone that could be either synthetic or natural corundum discloses no inclusions whatever, and curved striae or color banding are also apparently absent, it may be necessary to modify the illumination to assist in the detection of these features. Sometimes, this can be accomplished by viewing the stone against a very diffused light source or by immersion in a liquid such as methylene iodide, which has an index fairly close to that of corundum, and by using a diffused-light background. If an immersion cell is employed that has a base plate of clear glass, either opal glass or a fine grained white paper placed between the bottom of the immersion cell and the light is an ideal background for the resolution of color banding or striae. Sometimes, this is facilitated by moving a sharp edge of the diffusing surface slowly across under the cell, because the sharp demarcation between the dark and diffused-light background seems to increase the visibility of curved striae.

 When attempting to identify rubies and sapphires, a frosted- glass container may work particularly well. In some cases, color banding or striae may be seen more readily when the index of the liquid is not very close to that of the gemstone. For example, bromoform sometimes seems to give better results for corundum than methylene iodide. Usually, however, the closer the liquid in index, the better. Figure shows a color-banded natural sapphire photographed in air (A) and while immersed in methylene iodide (B).

Immersion is seldom necessary to disclose the deep-green overgrowth of synthetic emerald over pale beryl in the Linde synthetic emerald, but it is a good test to use when in doubt. When immersed in a liquid of 1.55-1.56, the coating shows up strikingly as a narrow green band rimming the stone (Figure E ).

Many other characteristics may become visible when a stone is immersed. Tiny inclusions that are masked by facets may become obvious. Faceted or rough stones with frosted surfaces appear to become transparent, if immersed in a liquid of very similar R.I. Thus, transparent material such as quartz immersed in clove oil, beryl in bromoform, or pyrope in methylene iodide may be examined internally without difficulty, even if the surface is badly frosted.

Immersion techniques are also helpful in the surface is badly frosted. Immersion techniques are also helpful in the detection of assembled stones, of dye in cracks, and of uneven distribution of color.

Immersion may be of great value in conjunction with the polariscope to determine optic character. A stone of high R.I. is often difficult to test by polariscope; particularly is it difficult to obtain an interference figure on a faceted stone of high R.I. Immersion reduces the facet reflections and the bending of light as it emerges from the stone, enabling the tester not only to determine more readily whether the stone is singly or doubly refractive, but to resolve interference figures as an aid in identifying a difficult stone.

 Immersion Vessels

A variety of immersion vessels are available; some are elaborate and others are very simple (Figure B). Usually, it is entirely satisfactory to use a very simple one. A flat-bottomed dish with very low walls is excellent for most purposes, because tweezers may be used to advantage to hold the stone from the sides. With a high-walled vessel, it is necessary to hold the tweezers-almost vertically, which may prove difficult. When the stone holder is held at a high angle, the surface tension of the liquid bending up or down at the tweezers distorts the image the viewer sees. This is very annoying when one is trying to determine whether color banding is straight or curved; in fact, it often makes it impossible to determine. Under these circumstances, the stone usually is placed in the bottom of the container and the tweezers removed. To move it, tweezers are introduced occasionally to turn or adjust its position in the liquid. Usually, the kind of flat-bottomed, low-walled vessels called cover
dishes are satisfactory; they are available at chemical-supply houses. The Institute makes a metal-walled, strain-free glass-based immersion cell that fits both the Illuminator Polariscope and the Gemolite.

At times, low, transparent, flat-bottomed glass ash trays are satisfactory for the purpose at hand. They are likely to cause some distortion, because usually they are molded rather than polished, but for most purposes they are relatively satisfactory.

More elaborate immersion vessels may incorporate a centered stone holding device that is controlled from outside the walls of the vessel and geared to be turned to any angle. Thus, a tester seeking to obtain an interference figure may adjust the position and angle of the stone easily. The stone holder in a cell of this type utilized beeswax on the end of a pedestal as the stone holder. A so-called universal-motion cell is very expensive and not essential.

Another type, which is very handy, is a small glass sphere with a liquid-tight cap, plus a means for mounting the stone near the center of the sphere. Such a vessel can be moved about freely until the interference figure becomes visible in the polariscope.

One other vessel that is sometimes useful is the type of flat cold-cream jar made of translucent to semi-translucent white glass or plastic. Light passes through it readily, but it is so reduced in intensity and diffused that color banding is not overpowered. Banding seems to stand out very well in such a vessel.


Early assignments in the Colored-Stone Course and the first refractometer assignment in this course made it clear that the greater the difference between the R.I. of a gemstone and the medium with which it is in contact, the greater the bending of light at the surface that separates them and the smaller the critical angle. If a hessonite garnet is placed in water, its facet edges do not seem as sharp, and its brilliancy seems to have been reduced materially. If it is placed in carbon tetrachloride, the effect becomes more apparent. The distinctness with which the stone stands out from the background, which is called RELIEF, is still great. When it is placed in bromoform, the edges seem to blur somewhat, but its relief is still evident. When placed in methylene iodide, it disappears; it has no relief at all. The reason is that the R.I. of hessonite is often identical to that of the methylene iodide. The degree to which a gemstone stands out from the liquid in which it is placed is a measure of the difference between its R.I. and that of the liquid. This fact has been used for years by mineralogists in the identification of powdered samples of minerals. An example of this is well illustrated in Figure F. The immersion liquid is methylene iodide (R.I. 1.74). The stone showing strong relief is diamond (R.I. 2.42), and the one that is almost invisible (i.e., low relief) is spinel (R.I. 1.72).

Another related factor is used by the mineralogist. A transparent mineral grain in a liquid shows a light rim. If the tube of the micro-scope is raised when the grain is in focus, the light rim moves towards the medium with higher R.I. This bright line is called the Becke line, and the process is the Becke method for R.I. determination. The mineralogist changes the liquid until he reaches one in which fragments almost disappear and the relife has gone to almost zero. When this occurs, the light edge assumes a color. The liquids used by optical mineralogists usually are mixtures calibrated to gradations of .01 or .005. Using these, it is possible to measure indices with great accuracy. For final accuracy, it is necessary to determine the orientation of grains and to determine the indices corresponding to the various directions in the crystal. On a less exacting scale, the indices of cut stones may be measured when they are immersed in the different liquids on the list given earlier in the assignment.

S.W. Anderson, of the London Laboratory, noted that when faceted stones were immersed in a liquid, those with a higher index than the liquid had a dark rim and the centers of the stones were light, the facet edges appearing as light streaks. Stones lower in index than the liquid had bright edges, and facet edges within the stone's image appeared black. How much higher or lower could be gauged by the width of the rim and the concentration of light or dark at the center of the stone. When there is nearly a perfect match between the stone and liquid indices, the edges will appear colored. If a record is desired, photographs can be taken by placing the cell with the stones immersed in it over a piece of photographic printing paper (slow film) in a dark room and exposing them for a second or two to light from an overhead lamp. The negative gives a clear idea of the index of the stones in relation to the liquid. For this test, it is best to use a fairly bright light with a narrow beam. A very diffused light is much less effective.

Although it is best for immersion purposes to use several of the liquids mentioned earlier, even water is sometimes helpful. Supposing a very large piece of jewelry contains many diamonds and the tester wants to check quickly whether some of the diamonds have been replaced by stones of lower R.I. Immersion in methylene iodide is even better, for it will still leave the diamond standing out, whereas any imitation will virtually disappear. An idea can be obtained by immersing the stones in water, but the higher the R.I. of the liquid, the easier the separation.


From the definition: specific gravity is equal to the weight of a storm divide by the weight of an equal volume of water. The weight of a stone can be found readily by weighing on an accurate balance. The weight of an equal volume of water can be found by the simple principle discovered by Archimedes, which states that "if a body is totally immersed in water, it loses an amount of weight equal to the weight of the water it displaces." Hence, the weight of an equal volume of water is equal to the weight lost by the stone when immersed. Thus :

S.G= Weight in Air
Weight in Air - Weight in Water

In other words, the specific gravity of a stone is the number obtained by dividing its weight in air by its loss of weight when weight in water. This gives a simple method of finding the S.G. of a cut stone.

Steps to calculate S.G.

  1.  Weigh the stone in air; assume this is 4.2 carats.
  2.  Weight the stone in water; assume this is 1.0 carats.
  3.  The difference in the two weighings gives the loss of weight in water : loss of weight = 4.2 - 3.0 = 1.2 carats.
  4.  S.G. of stone = weight in air divided by loss of weight in water : S.G. = 4.2/1.2 = 3.5.

(Note that the S.G. is a number that represents a RATIO, not a weight.)

For making S.G. determinations by the direct-weighing method, a delicate balance, such as a good diamond balance (Figure G.), is needed. Weighing the stone in air is simple, but to obtain an accurate weight in water is a slightly more complicated undertaking. The problem is how to weigh the stone in water with a minimum of time and equipment. The balance must be adapted so that the stone can be weighed accurately in air and in water. This is accomplished by using a few simple attachments, consisting of two thin wires, a glass beaker, and a suitable support (Figure E). These accessories can be constructed easily or purchased by those who have a diamond balance and who wish to add the necessary equipment.

The small metal support is bridged over one balance pan, so that it does not interfere with the free swing of the balance. The beaker of water is then placed on this stand and one of the thin wires is bent to form a basket at one end to hold the stone. This is suspended from one arm of the balance and immersed in the liquid. The other wire is suspended from the opposite arm of the balance and then clipped or filed off until it balances exactly the wire suspended in the water.

The surface tension of the water acting-on the wire at the point where it enters the water causes a damping effect (drag) on the free swing of the balance. This considerably reduces the accuracy normally obtained from the balance. However, any popular household detergent effectively reduces the surface tension of water, and the pinch or two necessary does not alter the density of the water sufficiently to require a correction factor. The proper temperature for the distilled water used in the direct-weighing method is, as previously mentioned, 4°C or about 40°F. However, using water at ordinary room temperature about 70°F. will not materially affect the calculation.

The following example will make the method clear. Given a a yellow stone to test:

  1.  Weight of the stone in air = 12.89 carats.
  2.  Place the stone in the wire cage, remove all bubbles, and make sure it is completely immersed. Now weigh the stone carefully in water. It weighs 9.67 carats.
  3.  Loss of weight in water = 12.89 - 9.67 = 3.22 carats
  4.  S.G. = 12.89/3.22 = 4.00.
  5.  Looking up this value on an S.G. table, we see that the stone is probably yellow sapphire (corundum),either natural or synthetic.

It will be noted in S.G. tables that normal values, as well as the most generally encountered variations, are listed. However, no tolerance is considered for mechanical errors in making determinations. There is a definite lower limit in the weight of the stone being tested to which diamond-balance determinations can be given any accuracy.

A colored mineral usually varies in specific gravity from one specimen to another: due particularly to slight variations in chemical composition. Therefore, several similar minerals may have the same S.C., because their values overlap. The values for specific gravity in all mineralogy texts and in some gemology texts are for all verities of a given mineral species. Such tables may add unnecessarily to the confusion of the gemologist who is using them and who needs to consider only gem qualities of the various species. The figures given throughout this course are for gem materials, and show specific gravity to be one of the more important tests.

To illustrate the effects of errors in weighing, refer again to the example given above. The weight of the stone in air is 12.89 carats and its weight in liquid is 9.67 carats. Suppose a mistake of one point had been made in weighing the stone in liquid. If the weight had been 9.66, instead of the correct 9.67 figure, the difference between 9.66 and 12.89 is 3.23 carats. By dividing 3.23 into 12.89, a specific-gravity determination of 3.99 is made. Therefore, an error of only .01 occurs because of a mistake of one hundredth of a carat in the weight. A stone that weighs 1.20 carats in air and 90 points in water would have an S.G. of 4.00. If, however, a mistake is made in the weighing and the weight in water is determined as .89 carat instead of .90, the error will be greater. Dividing 1.20 by .31 gives a resulting S.G. of 3.87; therefore, the error is .13. Carrying this example one step further, if a stone weighs .40 carat in air and .30 carat in water, it has a specific gravity of 4.00. An error in this case of one point, giving a water weight of .29, would result in the determination being 3.63 to 3.64. In other words, an error of more than .35 in S.C. would result.

From the discussion above, it can be seen that specific gravity determinations by the diamond-balance method for stones of less than one-half carat are not accurate, unless very careful weighings are made. No such limitation exists when heavy liquids are used.