570 



NATURE 



[December 30, 1920 



it reaches the wetted part. It is not sufVicient, 

 therefore, to interpose a liquid film between solid 

 faces to get lubrication ; indeed, as the experi- 

 ment proves, water increases the friction ; it is 

 an anti-lubricant for ordinary faces of glass. 



Is, then, the quality of lubricant a property of 

 a fluid? Does water fail to act merely because it 

 does not possess that property to which the name 

 "oiliness" is sometimes given? .\nother simple 

 experiment supplies the answer. Instead of a 

 glass plate, let us use a plate of ebonite. The 

 glass plate does not readily slip on this. The 

 angle at which slipping occurs is steeper than 

 when a glass plate is used. Now, when the lower 

 half of the ebonite plate is wetted, it is found 

 that a glass bottle encounters relatively high fric- 

 tion on the unwetted part, but slips quite freely 

 on the wetted part. Water, in short, is an ad- 

 mirable lubricant for glass on ebonite. Here is 

 another plate, picked up at random in the labora- 

 tory of the Royal Institution. Its composition is 

 unknown. Tested in the same way, water has no 

 detectable influence on the friction between glass 

 and the surface of this plate. 



It will be well to confess at once that these 

 simple experiments raise questions which are as 

 yet without an answer, and that much of what 

 follows concerning them is merely tentative. 

 They seem to establish two things, the first being 

 the curious paradox that a film of fluid introduced 

 between two surfaces does not always decrease 

 friction — it may, indeed, very much increase it. 

 The second is that the quality of "oiliness " — the 

 quality, that is, which enables a substance to act 

 as a lubricant — seems to be not the property of 

 a given fluid, but only of that fluid considered in 

 reference to a particular surface. 



It is necessary at this stage to clear away a 

 possible explanation of the paradox. When two 

 solid faces are separated by a thin film of fluid, 

 capillary forces operate, and, in certain cases at 

 any rate, these forces act so as to resist slipping. 

 They will so act, for instance, when the movement 

 of one face past the other increases the area 

 of the free surface of the film. Water has a high 

 surface-tension ; the capillary forces to which it 

 gives rise are usually large ; therefore it is per- 

 tinent to ask whether, when a layer of water dim- 

 inishes the facility for the slipping of glass on 

 glass, it is owing to capillary action. A qualita- 

 tive answer is to be found in the fact that water 

 does in some cases, as when glass is applied to 

 ebonite, increase the facility for slipping ; and 

 the late Lord Rayleigh furnished the quantita- 

 tive answer. He calculated the magnitude 

 of the capillary effect and found it negligible com- 

 pared with thq actual friction of glass on glass 

 wetted with water. An appeal to capillary forces 

 of this type will not solve the paradox. 



Some light is thrown upon it when we inquire 

 into the state of the surface of glass that has its 

 friction increased by water. Surfaces of glass 

 "cleaned" in the ordinary way by rubbing with a 

 glass cloth, or glass faces which have been simply 

 exposed to the air, are in point of fact not clean, 

 NO. 2670, VOL. 106] 



but highly lubricated with a film of matter 

 derived from the cloth or condensed from the 

 atmosphere. This "grease" film is of invisible 

 thinness. It is probably of the order of i /i/i 

 in thickness — that is to say, one-millionth of a 

 millimetre. It can be removed by soap and water, 

 which in turn must be removed by a stream of 

 water, ' and the plates dried in clean air out of 

 contact with solids. The film reforms quickly — 

 very quickly in London air, and less quickly in 

 the country. A "grease" film also creeps over 

 a cleaned glass face from ordinary solids with 

 which it may be in contact. Still, when due pre- 

 cautions are taken— and they are many — it is pos- 

 sible to get a glass face which seems to be really 

 clean. 



The first property of clean faces is that their 

 friction, one for the other, is very high ; indeed, 

 it is impossible to make them slip past one another. 

 One glass plate may be forced past another, but 

 true slipping does not take place ; they tear at the 

 point or points of contact. It is easier, in short, 

 to disrupt the actual substance of the glass itself 

 than to get the surfaces to slip over one another. 

 Clean glass faces " seize " when they touch. 



When chemical substances are tested as lubri- 

 cants on clean glass faces, a remarkable fact 

 emerges — namely, that some are quite neutral in 

 that they do not alter the resistance to slip in the 

 least ; such are water, alcohol, benzene, and strong 

 ammonia. Other substances have some lubricat- 

 ing action, great or small — that is to say, they 

 decrease the force needed to produce slipping ; 

 such are the alkalis, trimethylamine 'and tripro- 

 pylamine, the fatty acids — e.g. acetic acid — and 

 the paraffins. Those fluids which act as lubri- 

 cants are not necessarily fluids of any considerable 

 viscosity ; indeed, a high viscosity is compatible 

 with the absence of any lubricating action other 

 than flotation. Thus glycerine facilitates the 

 slipping of clean glass on clean glass only when 

 it is present in quantity sufTicient to float the sur- 

 faces apart. On the other hand, acetic acid and 

 tripropylamine — substances of low viscosity — are 

 admirable lubricants of glass. 



None of the fluids tested was found to raise the 

 friction of clean glass faces. They were either 

 neutral, or decreased friction to a greater or less 

 extent. The power of increasing the friction 

 of glass faces which neutral fluids, such as 

 water, possess is due, not to their action on the 

 glass itself, but to the fact that they interfere 

 with the action of the invisible grease film. Water 

 on an ordinary glass face acts as an anti-lubricant ; 

 on really clean glass it is "neutral." 



All solid faces, however, do not distinguish 

 chemical substances into those which are 

 " neutral " and those which possess lubricating 

 properties. Nearly one hundred substances have 

 been tested on burnished faces of bismuth, and 

 in every case some decrease of friction was ob- 

 served. 



A comparison of the lubricating action of simple 

 chemical substances on clean faces of glass or 

 of bismuth would seem to show that the quality 



