We noted on page 184 that the DC spikes produced by recirculating New York City tap water in our peristaltic pump approximated 1000 to 1500 mv, as compared with only 1 to 2 mv for the human ECG. We also noted that on Trace 12 (Fig. 132, p. 186) the DC pulses produced by inverse electroendosmosis could be "salted out" by an electrolyte.

In order to examine these aspects in greater detail, our peristaltic pump was again fitted with a 1/4" rubber tube as shown on Fig. 131a. Three electrolytes (aluminum, calcium, and sodium chloride, respectively) were successively and incrementally added to the pump "reservoir" on a millimol basis. This naturally represented 3:1, 2:1 and 1:1 types of electrolytes. Leads from the oscilloscope were run to points G and 3 on the pumping system. The pump was operated at low speed, and the peak DC voltage produced by the recirculating electrolyte was recorded for each incremental addition of reagent. Results are shown as curves A, B and C of Fig. 136. We also replotted on this sheet the curve from Fig. 20 (p. 24) showing the dispersion of a dilute suspension of Minusil by sodium chloride.

It is evident from Curve C (Fig. 136) that a DC peak of 1400 mv was produced at a concentration of 2.3 millimols of NaCl. Therefore, it is reasonable that New York City tap water would produce DC spikes of this or greater magnitude. Takeoffs from Curves A, B and C were made at selected voltage outputs, with results as follows:

TABLE NO. 21
Reagent	AlCl3	CaCl2	NaCl
Type of Electrolyte	3:1	2:1	1:1	( Ratio )
Millimols @ 1400 mv	.38	.95	2.3	1.0:2.5:6.0
Millimols @ 1000 mv	.67	1.7	4.0	1.0:2.5:6.0
Millimols @ 600 mv	1.40	3.6	8.0	1.0:2.6:5.7
Millimols @ 200 mv	6.0	15	32	1.0.2.5:5.4
			Average  1.0:2.5:5.8

In Curve D of Fig. 136, it would seem that the Minusil was being "salted out" at values above 11 millimols/liter of NaCl. This decrease in Zeta Potential is of course closely associated with the compression of the double layer. Also, it would seem that the electrolyte at the "wall of the tube" was having its double layer compressed by these same forces of high ionic concentration. This, too, resulted in a continuous lowering of DC voltage produced by inverse electroendosmosis. These aspects raise pertinent questions:

a) What (in detail) happened to Curve D at 0.9 millimols concentration of NaCl to cause it to level off? And what happened at 11 millimols, where the curve abruptly reversed and then continued without change of slope until the colloid "salted out"? *
[ * Certain aspects of this problem have previously been reviewed in Chapter 8, pp. 25-28. ]

b) Why didn't the inverse electroendosmosis curve for sodium chloride (C) also change shape at around 1000 mv of DC pulse generation? And why didn't the 2:1 and 3:1 electrolytes also show such change? All these curves continued unabated to 2000 mv, or more.

c) Why did the 2:1 electrolyte produce the same mv of DC spike with about one–third the concentration of the 1:1 electrolyte? And why did the 3:1 electrolyte produce similar results with about one–sixth the (millimols) concentration of the 1:1 electrolyte?

At this juncture, it would seem that inverse electroendosmosis is here producing an orderly but, in many respects, uninvestigated pattern.

Reagent concentration in millimoles/ liter

FIGURE 136

With regard to the average ratio of 1.0 : 2.5 : 5.8 found in Table 21, it should be remembered that the familiar "ionic strength function" of Lewis and Randall* is:
[ * See Abramson, Ref. 4–6, p. 30. See also, Lewis, G. N., and Randall, M.–J. Am. Chem. Soc.–43, 1112 (1921). ]

µ = 1/2 £ C_i Z_i²

in which: µ = ionic strength

C_i = molar concentration of each type of ion in the solution

Z_i = valence of each type of ion in the solution

The relative ionic strengths of the three aforementioned electrolytes (at unit concentration) are:

NaCl = 1 X 1² + 1 X 1² = 2 ÷ 2 = 1

CaCl₂ = 1 X 2² + 2 + 2 X 1² = 6 ÷ 2 = 3

AlCl₃ = 1 X 3² + 2 + 3 X 1² = 12 ÷ 2 = 6

Thus the experimental values of 1.0 : 2.5 : 5.8 check reasonably well with theoretical concepts of 1 : 3 : 6.

The theoretical considerations relative to the "thickness of the double layer," and the formula (1 ÷ k) for its determination, are given in most texts*
[ * See Refs. 4–3, 4–4, 4–6, 5–1, 5–2, 5–9. ]

and will not be repeated here. The straight lines of Fig. 137 are a plot of this basic formula for 3:1, 2:1, and 1:1 electrolytes in the concentration range of 0.1 to 1000 millimols. "Thickness" is in Angstroms.

Shown as curved lines are replots of Curves A, Band C of Fig. 136. Note the general agreement of theoretical concepts with experimental data. It should of course be noted that there is no real relationship between the mathematical scales chosen for the thickness of the double layer (plotted in Angstroms — y axis on the left), and inverse electroendosmosis as plotted in millivolts (y axis on the right). The respective scales and their locations were chosen as a matter of convenience in order to have one set of curves situated close to the other. However, there is no doubt that these curves strongly indicate that inverse electroendosmosis is closely related to the "thickness of the double layer."

It is of interest that here we have two quite different sets of physicochemical data, each intimately relating bulk and surface phases of a liquid–solids system. But it should also be emphasized that:

a.) The one set of data is based on theoretical considerations; the other, practical.

b.) The one set represents computed values; the other, a measurement of direct observations.

c.) The one set is based on particle charge, which reflects adsorbed ions; the other, on practical volts, which reflect electrons.

Thus we have another significant, readily demonstrated and easily measured relationship between the ion and the electron.

In considering these two systems, we should note another prime difference. The Minusil, in the process of being tracked in the electrophoresis cell, must be considered a relatively static system. This is because at its maximum electronegative state when suspended in an NaCl solution, its electrophoretic mobility would approximate only 4.0 microns/sec. per volt/cm.**
[ ** Note from Fig. 19 (p. 22) that the maximum electronegative ZP possible with Minusil and NaCl is –50 mv. This repre­sents an electrophoretic mobility at 25°C of 50 ÷ 12.85 = 4.0. Overbeek properly notes that due to the "relaxation effect," the electrophoretic velocities of colloids should be increased (by an appropriate factor) for Zeta Potential values greater than about 25 mv. For his tables and graphs, see Ref. 4–4, pp. 109–117. On pages 330–331 we present curves enabling ap­propriate corrections for the "time of relaxation" and, also, curves which take into account the 47(pi) – 67(pi) "adjustment." ]

At a concentration of about 10 millimols/liter (SC about 2000 micromhos), the voltage imposed on an electrophoresis cell during measurement of ZP would approximate 15V/cm. As a consequence, we have an electrophoretic mobility of the colloid during measurement of only 4 x 15 = 60 microns/sec. This rate is only one centimeter in three minutes. In sharp contrast, the velocity of the moving "finger" of the rotor of the H.R. Flow Inducer, or the tip of a laboratory stirrer with 4" paddle (each at 160 rpm), would approximate 850,000 microns/sec.*
[ * This is based on the tangential speed of a pump rotor having a diameter of about 4", which gives a circumference of about 1 ft. This, at 160 rpm, equals about 3 ft./sec., or say 850,000 microns/sec. These computations apply equally to a 4" diameter laboratory stirrer. We repeat that the DC spike is produced by liquid movement as well as liquid flow, and the DC spike will be produced — unabated — if the recirculating tube is entirely clamped off so that there can be no actual liquid flow. ]

It is therefore obvious that the mechanical movement obtained with either the pump or the laboratory stirrer — which in turn produces the liquid movement­would be 14,000*
[ * 850,000 ÷ 60 = 14,000. ]

times the velocity of Minusil particles at the time of cell electrophoresis. On this basis, one would have to consider cell electrophoresis a measurement of the colloid under relatively static conditions, whereas both the colloid and the liquid of substantially all stirred or recirculated systems must be considered highly dynamic. We stress this difference in velocities by pointing to an example. The fastest jet planes today travel about 2000 miles per hour. This is only 2000 ÷ 4 = 500 times as fast as a person can walk. We submit then, that the speed of the colloid during measurement of its Zeta Potential, and the velocity of liquid movement during measurement of inverse electroendosmosis, form a ratio so high that it is seldom encountered.

Therefore, we conclude that Curve D represents measurement under static conditions; and Curve C, measurement under highly dynamic conditions. These conclusions raise the following questions, some of which we will answer. Others will have to await much further investigation and study.

Electrolyte Conc. in millimoles/liter

FIGURE 137

1.) Does an electrolyte or polyelectrolyte adsorb readily on a colloid (as in Curve D) at low ionic concentration, but only when there is very little or no movement (agitation) of the colloid? That is to say, when the liquid is moving (as in Curve C), does this retard adsorption at low ionic concentration?

2.) Does significant adsorption in a dynamic system occur only after ionic strength is built up to a moderately high degree — for Curve D, say 11 or more millimols/liter?

3.) Is this another way of saying that low bulk–stress will result in adsorption in a system where there is little or no movement of solids, but that high bulk–stress will be required to force adsorption in a dynamic system where there is significant solids movement? Have we long overlooked this fine physicochemical point because the act of stirring* per se
[ * There is seldom need for a high degree of physicochemical dispersion in systems undergoing continuous mechanical agitation — such as certain industrial slurries during processing; and blood systems. ]

mechanically "disperses" the system? Is this why optimum coagulation of dilute colloid systems requires highly delicate mechanical agitation? The need for considerable extension of both theoretical and practical working concepts of dynamic systems seems apparent if these conditions are veritable; and we have reason to believe they are.

The writer has devised two significant and critical experiments which throw light on this subject. But before these are discussed, it is advisable to briefly present the basic and pictorial concept of Zeta Potential which has gradually evolved since Helmholtz. The writer quotes from one of his papers on Zeta Potential (1961) in Chemical Engineering magazine.*
[ * "Zeta Pot6ntial: New Tool for Water Treatment" — Chemical Engineering magazine, June 26 and July 10, 1961. This paper also appeared in the Journal of American Water Works Assoc., Vol. 53, No. 8, August 1961, under the title of "Zeta Potential and Its Application to Difficult Waters." The writer wishes to acknowledge the kind, invaluable assistance of Dr. Karol Mysels in the preparation of the original paper which set forth these simple but basic concepts. ]

The fundamental principle of electro–osmosis (the forerunner and in many respects the inverse of electrophoresis) was discovered about 150 years ago.

In early experiments, it was noted that if a ball of clay was submerged in water, and two glass tubes were inserted in the clay with their ends projecting above the water, then a d.c. voltage connected to electrodes in the tubes would cause the liquid to be elevated above the water surface in one tube, and depressed in the other. A reversal in polarity would reverse the order of liquid elevation or depression. Thus, an immovable solid (the packed clay) causes a movement of the liquid. This is electro–osmosis.

In electrophoresis, a suspended solid (such as a discrete particle of colloidal clay) moves in a d.c. field in relation to the suspending liquid (water).

The basic meaning of Zeta Potential is set forth quite simply above. The pictured colloidal particle (Fig. 138) is ... electronegative, since it has adsorbed an excess of negative ions at its outer surface. The preferential adsorption (or existence) of negative ions at the surface of most colloids has been the subject of much investigation. An oversimplified explanation is that these are more often OH ions — and that they probably fit better into the lattice structure of the colloid, or are more acceptable to the colloid than the positive ions, which are often larger than the negative (ions)... .

Figure 138

The negative surfaces of the colloid attract a surrounding layer of positive ions, which may originate either from the "bulk of the suspending liquid", or from the surface of the colloid itself. The oppositely charged ions, or "counterions," are drawn to the colloid by electrostatic attraction, while thermal agitation or Brownian motion tends to distribute them uniformly throughout the solution. This charged system — the surface of the colloid and the neutralizing counter–ions — is called a "double layer."

If the negative charge is large, some counterions will be so strongly attached that they will stick to the surface of the colloid as a firmly attached compact layer, often called the Stern layer. This layer partially neutralizes the charge and electrostatic attraction of the colloid so that the remaining counter–ions can on the average be farther away, while still being kept in the immediate vicinity of the colloid. They form the so–called diffuse part of the double layer.

The attraction of the central colloid is greatest, of course, close to itself, both because of the distance involved and also because the counter–ions near it interpose their positive charge and thus shield those counter–ions farther away. Hence, the neutralizing counter–ions are most concentrated near the colloid and become gradually negligible farther away. Similarly, negative ions of any salts present tend to be repelled from the immediate vicinity of the colloidal particles. How far away this diffuse double layer extends depends primarily on the concentration of simple salts in the liquid; if their concentration is large, this distance may be reduced to a few Angstroms.

The Zeta Potential is the potential at the surface separating the immobile part of the double layer from the diffuse part. It is a simultaneous measure of the charge of the diffuse double layer (per unit surface of the colloid) and of its extent away from this surface. The Zeta Potential is therefore related to the force and distance over which the particles can repel each other and thus prevent flocculation.

This, of course, is considerably oversimplified. Verwey calls attention to the fact that the surface charge on the colloid may reflect the nature of the adsorbed capillary–active ions, or the nature of the potential–determining ions (principally valency or equivalent valency) situated in the double layer, or both.

The Schulze–Hardy rule states that in the coagulation of sols, the flocculating (hence Zeta Potential controlling) power of bivalent ions is approximately 20 to 80 times that of univalent ions — and the flocculating power of trivalent ions is 10 to 100 times that of bivalent ions.*
[ * See Table 43, p. 415, Ref. 5–2 — Rutgers, Physical Chemistry. ]

When a liquid containing such charged particles is placed in an electric field, the negative particles are attracted to the positive electrode, and the counter–ions to the negative. This attraction increases with the charge on the particle. Friction between the particle and the surrounding liquid containing the diffuse double layer slows down the resulting motion toward the electrode — the greater the extent of the double layer, the lower this resistance. Therefore, particle velocity in a given field increases with both densitv and extent of the double layer, which, as we have just seen, are measured by the Zeta Potential. Hence, the velocity of a colloidal particle in an electric field is proportional to that field (volts/cm) and to the Zeta Potential of the particle.

Now again consider Fig. 138 (p. 199). This is a pictorialization of the conventional concept of Zeta Potential. It suffers from being planar rather than three–dimensional, and from inability of a draughts–man to adequately picture the number and relative density of ions surrounding the particle.*
[ * It is of course impossible to draw a sphere in three dimensions, except through air brush (or equivalent) techniques. ]

However, the two bottom curves of Fig. 138 (which are shown enlarged on Fig. 141, p. 203) enable one to visualize the preponderance, plus the relative concentration of positive ions at the plane of shear surrounding the particle. Note that both the preponderance and the concentration diminish to a state of ionic balance at the start of the "bulk" phase.

Figure 141

The writer knows of no text which states that this model obtains for either a static or a dynamic system. The general inference has always been that it refers and applies to both. But it would now seem that it is much more appropriate for a static colloid system, and we reiterate that a suitable extension of concept must eventually be made to better accommodate it to dynamic conditions. In the meantime, this concept must of course be employed for both states. In most practical situations, the thing that is principally gained is a better concept of the basic factors controlling colloid stability.

We relate the terms static and dynamic to the colloid only, because the ions and molecules of a system are never static — but always in a highly dynamic state of Brownian Motion.**
[ ** Perhaps the best visual example of Brownian Motion extant is that of cigarette smoke confined in a small glass ring positioned under the stereoscopic microscope, as shown on Fig. 2 (p. 5). It must be seen to be appreciated. ]

Indeed, their exact position at any moment can be considered only on a statistical basis. But, for that matter, we cannot draw any hard and fast rules, because the visible colloid of 0.3 to 1.0 microns is also undergoing a degree of Brownian Motion. It seems hardly possible that we can ever take these many imponderables into full account. But we can profit greatly by a clear understanding of these factors. They play a leading role in such divergent areas of colloid stability as paint manufacture — and blood coagulation.

Refer again to Fig. 138 (p. 199) and visualize three types of dilute colloid systems consisting of 1000 ppm of Minusil in distilled water — with or without electrolytes as indicated. Bear in mind that Minusil in a "neutral" unstressed system has a Zeta Potential of –30 mv. First visualize the system as static, with no mechanical movement; and then under a sta te of moderate agitation.

Case  1: The colloid suspended in distilled water at its "natural" Zeta Potential.

To stress the point, let us assume that the water was ultra–pure, with no dissociation.*
[ * This, of course, is impossible from a practical point of view. ]

Here the thickness of the double layer would be so large as to be almost infinite. The ZP of the static system would approximate –30 mv.

If the system is then placed under agitation, the ZP will remain unchanged, because the colloid has virtually no double layer of counterions from which to be forcibly dislodged by mechanical movement. Hence, the dynamic ZP would remain at –30 mv.

Before examining the next case, it seems advisable to consider spatial relationships. Actually, there is no pure water. A fresh, properly distilled laboratory sample will run about 2 micromhos of conductivity, equivalent to say 1 ppm of NaCl. Dissociation is due principally to carbonic acid, and this low concentration of H₂CO₃ would approximate 0.01 millimol/liter.

Fig. 139 shows the graphical solution of the "Standard" formula (1 ÷ k) for the "thickness of the double layer." In essence, this is the same plot as the straight lines of Fig. 137 (p. 197) except that six, rather than four types of electrolytes are shown; and five, rather than four cycles. Also, the formula is listed at the top of the graph.

Remember, it is the double layer that promotes colloid stability. When a system is stirred, it is a moot question how well the counterions can keep pace with the moving colloids.

ppm NaCl — for reference

FIGURE 139

Case 2: The colloid suspended in a highly disperse system-at high electronegative Zeta Potential.

We now consider the system dispersed by sodium tetrapyrophosphate, a 1:4 anionic electrolyte shown on Fig. 19 (p. 22). Here, at 1.0 millimol, ZP would be at its maximum electronegative value of –70 mv, and from Fig. 139 the thickness of the double laver would approximate 30 Angstroms.

The strong electronegative Zeta Potential of the colloid is due solely to its surrounding ionic cloud — its diffuse layer. If the colloid is agitated, i.e., quickly "pushed outside" its double layer — or moved so that its double layer becomes eccentric rather than concentric — or moved at a rate faster than its ionic envelope can keep pace (see Fig. 140, p. 202), then its Zeta Potential must drop. The exact extent of this drop is presently (and probably will remain) unknown. Thus the ZP of a static system would be maximum, and the ZP of the same system under dynamic conditions would be a lesser value. Hence, with agitation this system would become less stable.

Case 3: The colloid suspended in a highly coagulated system — at zero Zeta Potential.

Maximum coagulation occurs at zero Zeta Potential, and we can achieve it with say aluminum chloride, a 3:1 electrolyte shown on Fig. 19 (p. 22). Obviously, zero ZP is achieved through neutralizing the net excess electronegative charge on the Minusil by the trivalent aluminum ions. In this instance, there is no chemisorption. When the charge is neutralized and mutual repulsion is eliminated, agglomeration results. This is due to the attractive Van der Waals' forces — which then become effective. Thus, the static state of Case 3 is characterized by zero Zeta Potential and massive agglomeration.

Now when mechanical movement is created, the colloid is "pushed out" of its concentric protective field of electropositive counterions. Resultantly, its Zeta Potential must immediately increase to some electronegative point between 0 and –130 mv (more probably from zero to about –10 or –115 mv).

FIGURE 140

We now digress to briefly note that from a practical (rather than theoretical or mathematical) point of view, the thickness of the double layer represents the distance from the plane of shear*
[ * Some will want to argue the exact point of measurement. ]

surrounding the colloid, to the approximate centroid of the area graphically shown on Fig. 141. (See also Fig. 138, p. 199.) This area is bound on one side by the curve representing the concentration of positive ions, and on the other side by the curve delineating the concentration of negative ions. In many respects it portrays the ionic cloud surrounding the particle. The exact extent of the diffused layer is mathematically indeterminate, in that its net charge asymptotically approaches zero with increasing distance from the plane of shear. From a practical standpoint, the thickness of the diffuse layer may be assumed to be about three times the computed "thickness" of the double layer. It is a reasonable approximation for most situations.**
[ ** See Ref. 5–9, pp. 251–273; also Refs. 4–3 and 4–4. ]

Figure 142 shows the thickness of the diffuse layer, which is essentially a replot of Fig. 139, with each ordinate multiplied by a value of 3. That is to say, these curves give values three times as large as those of the conventional "double layer."

The average point–to–point spacing of ions for any given molar concentration is of interest. The geometric arrangement which will make each ion equidistant from its neighbor is a tetrahedron, and the formula for expressing this distance is:

d = { (12 ÷ 5 X [the square root of 2] ) X (1 ÷ 6.02) X (10⁷ ÷ mv) }^1/3

If one considered the spacial arrangement as simple cubic, then the formula would be:

d = {10⁷ ÷ (6.02 X mv) }^1/3

Notations are as follows: d = distance, center to center of ions, in Angstrom units.

m = concentration of the reagent in millimols per liter.

v = the sum of the valences of the anion and the cation.

The cubic formula may be converted to the tetrahedral by multiplying by a factor of 1.19. Conversely, to convert tetrahedral to simple cubic spacing, the factor would be 0.84.

Figure 139 (p. 201) shows that a concentration of 1 ppm of NaCl is equivalent to 0.018 millimols. The thickness of the double layer under this condition would be about 720 Angstroms. Since Minusil has a radius of about 5,000 Angstroms, the double layer would represent an ionic cloud or envelope with a thickness of 720 x 100 ÷ 5000 = 14% of the colloid radius. Thus for colloids of 1 micron size, the thickness of the double layer (ranging from about 1 to 1000 Angstroms) would approximate 0.06 to 15% of the radius of this colloid.

Figure 20 (p. 24) shows that most colloids in dilute suspension are near maximum dispersion at an electrolyte concentration of about 2 millimols/liter, giving a thickness of the double layer approximating 70 Angstroms for a 1:1 electrolyte, and 30 Angstroms for a 1:3 electrolyte. In a practical sense, this means that if a colloid the size of Minusil is moved very quickly a distance of 2% of its radius (i.e., 100 Angstroms), and if its diffuse layer of counterions cannot "keep pace," then it has moved to and beyond the limit of its ionic cloud. It has temporarily become eccentric, instead of concentric, with its surrounding counterions (see Fig. 140). This is another way of saying that a 1 micron colloid, if moved quickly, does not have to be moved very far to have its position in the double layer materially affected. We reiterate that the thickness of the diffuse layer is three times the values listed above.

Figure 140 shows a concept of this spacial relationship, i.e., the colloid being quickly forced out of concentricity (with its double and diffuse layer) by mechanical movement. These figures are not to scale. The diffuse layer is shown much "thicker" than it actually is. To scale, the double layer (at say 1% of the radius of the colloid) would appear as a circle, about the thickness of a pencil–line.

Figure 141

Figure 142

Here then, we have a basic principle which will cause aqueous colloid systems to behave as follows:

Case 1. Agitate a "natural" system, and there is no change.

Case 2. Agitate a highly disperse system, and it becomes less disperse.

Case 3. Agitate a highly coagulated system, and it becomes more disperse.

Now discard Case 1, and consider Cases 2 and 3 as highly concentrated slurries, instead of dilute systems:

Case 2. Shake the liquid slurry, and it turns to a semi–solid.

Case 3. Shake the highly coagulated gel–like mass, and it turns to a liquid.

This brings us to the real crux, for we believe these systems actually represent the true physico–chemical principles of:

Case 2. Dilatancy

Case 3. Thixotropy

We now reverse the order slightly and discard Case 2.

Case 3. Keep the slurry moving, and it remains in a liquid state. Stop its movement, and it coagulates.

[ A good example of this last example is concrete. When it is in the truck continuously being agitated, it will remain liquid for a considerable length of time. Once it is removed and allowed to be still, it will set in a predetermined time determined by its Zeta Potential. The Zeta Potential is engineered to give concrete workers enough time to get it in place, and then it quickly sets, allowing them to be able to "surface" it in a timely manner. — Tommy C ]

No one can fail to recognize this last sequence as a replica of the action of certain industrial colloid systems such as paint-and blood, when shed.*
[ * The writer does not believe that thixotropy, or the concept that motion disturbs the double layer and thereby reduces the stability of a dispersed system, has ever been employed in blood work. However, the concept that lack of motion (of blood) is associated with coagulation would seem to date back at least to Aristotle (about 350 B.C.). Harvey (1628) in his Motion of the Heart and Blood in Animals (Ref. 12–34, p. 94) states: "The blood, therefore, required to have motion, and indeed such a motion that it should return again to the heart; for sent to the external parts of the body far from its fountain, as Aristotle says, and without motion, it would become congealed." Harvey knew that disease induced coagulation of the blood, for he states (p. 97): "... tartan fever ... renders the patient short–winded ... and indisposed to exertion because the vital principle is oppressed and the blood forced into the lungs and rendered thick."

In studying Harvey's paper on circulation, one is constantly impressed by the fact that this is one of the most straight–forward and perceptive pieces of technical writing in existence. And his entire concept and expression is within the grasp of even the most modestly educated person. One wonders why the medical profession takes such a dim view of new scientific concepts — even those which arise from within its ranks. When Harvey presented his thesis on circulation, he opined (p. 70): "But what remains to be said ... is of a character so novel ... that I not only fear injury to myself from the envy of a few, but I tremble lest I have mankind at large for my enemies ... ." When Jenner presented his findings concerning vaccination and smallpox, the president of the Royal Society admonished that he "should be cautious and prudent ... and ought not to risk his reputation by presenting to the learned body anything which appeared so much at variance with established knowledge ... ." By contrast, schoolboys (uninhibited by the Hippocratic oath) were successful in building short–wave sets and engaging in world–wide radio transmission within a very few years after DeForest announced his vacuum tube. ]

Here we have (though not through happenstance) a principle that in many respects can be regarded as the model for an "ideal" blood system — a system that is capable of handling a high suspended solids concentration — and one that tends to remain fluid as long as the system remains in motion. But if motion ceases and the system changes from dynamic to static, it tends to coagulate, gel, or rigidify. Our use of the expression "tends" denotes a "Permissive" rather than a "compulsory" physicochemical state. By permissive, we imply that this reaction is inclined to and capable of going through to completion — provided certain other conditions also obtain.**
[ ** This of course is the very essence of catalytic activity, which is widely employed despite the fact that it is sometimes not well understood. There also can be chemical reactions which are simultaneously dependent upon the presence of three or four reagents, with none of them performing a catalytic function. ]

Homeostasis connotes the maintenance of constant, appropriate, conditions of the body's interior — as Claude Bernard so aptly pointed out. The circulation of blood is among the most important of these. For even the most primitive life to come into existence, nature had to devise a blood system which was "self–sealing." That is to say, blood had to remain fluid in situ to permit its recycling, but in some long–unknown fashion it also had to possess the ability to seal off all minor leaks occasioned by normal trauma, or "when shed." This capability is as old as biology itself, and it emphasizes that the basic laws governing blood coagulation cannot be the result of evolution, for in essence they enabled evolution. Thus, to postulate a physicochemical principle which tends to promote dynamic fluidity and static rigidity is to postulate a principle that had to exist — even before biological systems could come into being. And, as Max Planck said, a law of this universality must be characterized by stark simplicity. No principle tending to hold dynamic systems fluid, or promoting coagulation of static ones, can be simpler and more fundamental than the ones outlined in Cases 1, 2 and 3.

We now describe the procedures for preparing colloid systems whose stability is delicately and dramatically controlled by Dilatancy or Thixotropy. They are quite simple, and we will first deal with Dilatancy.

We noted on pages 29-31 that the end point of "spatula fluidity" for Minusil was about 64%. That is to say, a "natural"*
[ * We use the term "natural" to denote a colloid system or slurry made up simply with distilled water, and containing no applied electrolytes or surfactants. ]

slurry up to 64% suspended solids (wt/wt basis) will be sufficiently fluid to permit a vertically held spatula to be drawn through the system without leaving a "track." Above 64%, the system rapidly becomes paste–like and Stiff, unless an effective dispersing agent is added. Remember that the ZP of a Minusil system remains substantially constant at –30 mv throughout the entire range of 10 to 640,000 ppm, due to balanced bulk–stress. Now if we add the proper amount of an anionic dispersing agent to force the ZP of a Minusil slurry to its maximum electronegative state, say –60 mv, maximum fluidity will also be achieved. Moreover, at this point it is possible to bring the suspended solids content to maximum. In this highly disperse state, each colloid resists its neighbor with sufficient force to create a degree of mutual repulsion that acts somewhat as a "lubricant." One could compare this action to that of a "Hydrofoil" (a boat which "rides" on a cushion of air); or to the action of two like poles of bar magnets when brought close together.

Now if we bring an 80% Minusil slurry to a ZP of –60 mv, it will be quite fluid. But this same suspended solids concentration would be a harsh paste at –50 mv. Therefore, if we prepare a fluid slurry at –60 mv, and agitate it to a point where its ionic envelopes are "deformed" — causing the ZP to lower to –50 mv, the slurry will immediately become a stiff paste.

A highly disperse Minusil slurry of 80% suspended solids will have the consistency of heavy cream.**
[ ** For best results it is recommended that the slurry be freshly prepared each day. If it setties, "close packing" renders it useless. ]

But it can be poured slowly into a W ng blender with no difficulty. When the blender is started, the rotor will instantly bind and, unless quickly relieved, the motor will burn out. This is a dramatic and convincing experiment, and of course the slurry is a conversation piece. Another revealing experiment is to attempt to thrust a metal rod into a 600 ml wide–mouthed bottle of the slurry. Although under slow stirring the slu y will remain highly fluid, this quick "jab" creates the sensation of plunging the rod into a thick pad of sponge rubber. Quick withdrawal of the rod from the jar is likewise difficult, and the colloid mass which momentarily clings to the rod resembles a stiff putty. Employing sodium hexametaphosphate as the dispersing agent, the quantities are readily computed from our previously developed formula on pages 35-36.

Q = ax + by = 0.090a + 2.3b

Set b = 800 (80%), which automatically makes

 a = 200

Q = 200 X 0.090 + 800 X 2.3

    = 18 + 1840

    = 1858 mg = 1.858 g.

Procedure:  Dissolve 1.86 grams of sodium hexametaphosphate in 180 ml of water, and make up to 200 ml. Slowly add 800 grams of Minusil — mechanically agitating as long as possible — then stirring lightly and very slowly by hand. The resulting slurry will exhibit Dilatancy to a high degree. The concentration can sometimes be brought to even 82 or 85%.

We now state that our basic formula ( Q = ax + by ) is as applicable for coagulation as dispersion. The formula for Minusil and aluminum chloride*
[ * This formula will be developed in the next chapter. ]

is ( Q = 0.0007a + 0.50b ). For a 60% thixotropic suspension of Minusil:

b = 600 (60%);   therefore, a = 400

Q = ax + by

    = 400 x 0.00070 + 600 x 0.50

    = 0.3 + 300 = 300 mg = 0.30 grams

Therefore to prepare a 60% thixotropic suspension, dissolve 0.30 grams of aluminum chloride (AlCl₃ • 6H₂O) in 350 ml of distilled water and make up to 400 ml. Then add 600g Minusil, agitating as required.

If this thixotropic slurry is placed in say a 600 ml Griffin tall form beaker, an examination will reveal it to be quite viscous and gel–like. If the beaker is subjected to a strong rotary motion, the colloid will be "thrown" out of its ionic envelope; the Zeta Potential will become somewhat electronegative and, resultantly, the system will become disperse and quite fluid, Cease rotating the beaker, and the slurry will immediately return to a thick, gel–like paste.

A physicochemical sequence such as this, which causes a concentrated slurry to remain fluid during circulation — but to gel when motion ceases, represents a situation which calls for due pause. It cannot be considered mere chance. These are simply manifestations of a natural law that now becomes more understandable. But there are many natural laws about which we have no understanding.

The writer has felt for many years that the elusive principles of Dilatancy and Thixotropy had to be elucidated before one could hope to draw human blood which would not (naturally) coagulate; that is to say, without the use of any anti–coagulant, in vivo or in vitro.

Admittedly, our demonstrations herein of Dilatancy and Thixotropy were with Minusil and electrolytes — not with a system consisting of red blood cells, Platelets, electrolytes and the plasma proteins. But we believe that if basic principles are understood, one can change from Minusil to red cells, or colloids, or suspensoids, with few difficulties.

Moreover, several years ago we worked out another unknown "facet" of an entirely different (but highly important) nature, which we also consider a major principle in blood stability.

Additionally, and of considerable significance in coagulation, is the effect of long– and short–chain polymers upon the stability of the system. For many years, the writer has employed such polymers in his professional practice, and some years ago Healy and LaMer*
[ * See Ref. 6–10. ]

mathematically rationalized this phase of physical chemistry.

Therefore, in March 1967, armed with these three factors, plus ten years of basic research and practical experience in Zeta Potential which encompassed many hundreds of different industrial systems, we felt that we had sufficient "armamentarium" to make an attempt to draw blood which would not coagulate. We thought we might do so without triggering the delicate mechanisms which we (presently) think are responsible for its coagulation. We were well aware of the odds against success and, had we failed, we would have been neither surprised nor discouraged. We would simply have started to look for one more factor.

It does not seem logical that Nature would have entrusted the entire future of its biological systems to the proper functioning of any single principle designed to control blood stability. We believe that there had to be involved at least two or three (possibly more) different but interdependent major physicochemical principles; and Thixotropy can well be one of these.

We therefore requested the cooperation of the New York Chapter of the Red Cross in our attempt to perform this most important and extremely delicate exsanguination. We had, over the years, enjoyed a pleasant rapport with them. They had furnished us with both outdated and fresh blood for research, and they were aware of our continuing work on the physical chemistry of blood systems.

For this experiment, we proposed to furnish our own donor and all the necessary apparatus and personnel. We expected them to furnish only a technician for exsanguination, who would employ their normal facilities. The Director of the New York Chapter provisionally agreed to this, if we could obtain permission from the Research Division of the National Red Cross, with headquarters in Washington, D. C.

The date of April 14, 1967, at 10 A.M., was tentatively set. But as the date drew near, the obstacles they imposed became so numerous and onerous that we finally were forced to call the whole thing off.**
[ ** We do not mean to imply that it was literally impossible for us to work with the Red Cross. But we simply do not have the time, patience, or financial resources to cope with uphill situations such as this became. We respect the rules, regulations and prerogatives of others. But we naturally resent any effort to place unwieldy and, in our opinion, unnecessary impediments in our path. ]

This episode jarred the writer to a point of real frustration and brought to mind the expression by Abner Dean "WHAT AM I DOING HERE"?*
[ * This expression is the title of a book of drawings by Abner Dean, a realistic depictor of human mores. It was published by Simon and Schuster, New York, in 1947. ]

I believe I can be pardoned for mentioning the fact that for ten years our organization has borne the entire cost of continuous research on Zeta Potential. Much of this work has been purely scientific and in the public interest — not in the area of industrial science. Since the latter represents our professional activity, it should be quite understandable that lack of cooperation such as we experienced from the Red Cross greatly discourages our continuation of this work.

A friend, who is a very fine surgeon but somewhat of a skeptic, tells me it is far easier to remove a one–pound tumor, than 1 milligram of hypocrisy. My experience is beginning to reveal just how great a mass 1 milligram can be. It is also becoming increasingly apparent that the problems involved in the physicochemical aspects of cardiovascular disease are far simpler than those one encounters when dealing with human nature. Continuing this line of thought, we believe it timely and appropriate to include a photograph of a clay model done by our Joel Rudnik. Joel also selected the title.

Scientists sometimes question those who introduce new concepts!

FIGURE 142