16 Dr. P. E. Shaw .
unaltered , we raise the value of by 50 per cent. , so the question of the compressibility of the material is not unimportant .
The above result agrees with Sir Joseph Larmor 's ( p. 14 , supra ) and with that obtained by combining Terazawa 's equations ( 6 ) and ( 7).* I was not aware of the existence of Sir G- .
H. Darwin 's paper until long after my paper was published , and did not notice the reference in it to Lord Kelvin until the publication of Terazawa 's paper led me to restudy the problem .
Experiments on Tribo-Electricity .
I.\#151 ; The Tribo-Electric Series .
By P. E. Shaw , B.A. , D.Sc .
, University College , Nottingham .
( Communicated by Prof. E. H. Barton , F.R.S. Received May 24 , 1917 .
) CONTENTS .
PAGE .
I. Historical ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... 16 II .
Various Short Tribo-Electric Series ... ... ... ... ... ... ... ... .
18 III .
Apparatus and Methods ... ... ... ... ... ... ... ... ... ... ... ... ... 21 IV .
The Full Tribo-Electric Series ... ... ... ... ... ... ... ... ... .
23 V. Theory ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... 27 VI .
Conclusion ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... . .
31 VII .
Summary ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .
32 List of References ... ... ... ... ... ... ... ... ... ... ... ... ... ... 33 The term " Tribo-electricity " is used in 0 .
D. Chwolson 's ' Traite de Physique ' ( 1 ) .
This term is a convenient equivalent for frictional electricity ( Tpofirj = a rubbing ) .
I.\#151 ; Historical .
Little exact knowledge of tribo-electricity has yet been accumulated , and this subject has certainly not been raised to the dignity of a quantitative science .
In the present paper will be found an account of experiments in which the conditions of the solid bodies rubbed together have been greatly varied .
Thus , temperature has been changed both before and during friction ; the surfaces used have been rubbed with considerable pressure while hot ; they have , wherever possible , been rubbed together when flexed ; and they have been prepared before rubbing by being ground and polished in various ways .
Much information has thus come to light as to the .
electrical surface conditions of a variety of solids .
* 'Phil .
Trans. , ' A , vol. 217 , p. 37 .
Experiments on Tribo-Electricity .
Before setting forth the methods and results of this paper , let us note the information previously available .
Good summaries of the subject will be found in Wiedemann 's ' Electricitat ' ( 2 ) and Chwolson 's treatise ( 1 ) .
The data on the subject altogether lack coherency .
The electrical effects obtained by friction had been so often found to be irregular\#151 ; an irregularity which in great part I explain in this paper\#151 ; that the subject was in general avoided .
One obvious source of uncertainty of action is that rubbing necessarily affects , perhaps destroys , the surfaces under investigation , so that friction , especially if violent , may , in effect , give rise to surfaces of a new kind .
As to recent research on the subject , Hesehous ( 3 ) , in the opening years of this century ( 1901-5 ) , brought to light in , a series of papers , some new phenomena , and made an attempt to reduce the known facts to some sort of order .
He suggested certain generalisations as follows : ( 1 ) a polished surface is always + to a matt surface of the same material ; ( 2 ) a more dense surface is always + to a less dense one of the same material ; ( 3 ) dust proceeding from a body is always \#151 ; to the solid substance it leaves .
He also made investigations of the action of radium , and he found some materials , e.g. , glass , are made more + -forming , whereas others , e.g. , ebonite , are made more \#151 ; forming by the discharge .
Further , Hesehous attempted to bring tribo-electric phenomena within the scope of the electron theory .
He supposed that electrons pass from the more dense to the less dense of any two bodies brought into contact , this movement of the electrons being facilitated by the change , due to contact , in the surface tension of both bodies at the point of contact .
Of the above generalisations , the first and second are shown below to be incorrect ; the others have not been directly put to the test .
A few years ago some quantitative researches ( perhaps the first ever made on tribo-electricity ) came from Bangor , North Wales : from Morris-Owen in 1909 ( 4 ) , and from W. Morris-Jones in 1915 ( 5 ) .
The apparatus used in the two cases was of the same type .
It was arranged that not only the charge ( Q ) produced , but also the work ( W ) done in producing it , and the pressure ( P ) applied in the friction between the solids rubbed , could be actually measured throughout an experiment .
Simple relations were found connecting Q , W , and P. In 1915 I investigated the anomalous effects obtained from glass after heating ( 6 ) .
It was found that glass , which is + to silk , becomes temporarily \#151 ; to it after being raised to sufficiently high temperature .
Further , this anomalous effect was found in a great number of other solids , including metals and ebonite , if rubbed with suitable materials .
The general conclusion was reached that this behaviour , after heating , is common to all solids .
VOL. xciv.\#151 ; A. c Dr. P. E. Shaw .
The last-named research was the starting part of the present experiments .
Its methods have been developed and its conclusions confirmed , while the range of enquiry has been extended to several physical changes other than thermal .
II.\#151 ; * Various Short Tribo-electric Series .
In the early pages of most books on electricity is found a series of solid materials arranged in order in such wise that any one of the materials becomes -f if rubbed with any other lowrer down in the series .
The arrangement of materials by different authorities ranging over the last ] 50 years may be here contrasted .
In the last list I take from my Table ( p. 32 ) only such materials as are mentioned in the other lists .
Wilcke ( 1759 ) .
Faraday ( 1840 ) .
Jarnin and Bouty 's ' Physique ' ( 1891 ) .
Shaw ( 1917 ) .
G-lass Wool Quills Wood Paper Ground glass Pb Sulphur Metals Cat 's fur Wool Quills Flint glass Cotton Linen Silk Hand Wood Fe , Cu , Ag , Pb .
Sulphur Cat 's fur Glass Wool Feathers Wood Paper Silk Resin Ground glass Glass .
Wool .
Cat 's fur .
Pb .
Silk .
Paper .
Cotton .
Wood , Fe .
Ground glass .
Resin .
Cu , Ag .
Sulphur .
If lines were drawn joining like materials in the different lists , there .
would be many intersections of these lines , showing how differently the solids are arranged .
The many disagreements thus displayed show how necessary it is to standardise the materials used .
As will be shown in the sequel , it is probable that these dissimilarities arise from the use of different methods or of different materials , or of materials whose surfaces differ in physical condition , rather than from personal errors of observation .
This kind of research may seem easy : it is easy to get some results of sorts , but it requires considerable practice and special knowledge of surface conditions to obtain results which will stand repeated tests .
The series which I give in the Table ( p. 32 ) is the last of some dozen or so of lists drawn up during many months .
At first , errors were made through the employment of faulty methods or from ignorance of the many factors involved .
It took considerable time , as one 's skill increased , to sort out the materials into their correct places .
When the series was considered correct , I prepared a \ Experiments on Tribo-Electricity .
complete new set of surfaces , and checked the series by going over the whole ground again with these fresh materials .
My complete series is shown in the Table ( p. 32 ) .
This includes only fairly common solids and one liquid , mercury ( whose peculiar action is described later , p. 25 ) .
It is desirable to have a long list , such as this , which contains about seventy-eight materials , for , by including many materials , we obtain more steps from extreme + at the top to extreme \#151 ; at the bottom .
If the steps are many , the gradations will in general be less , and thus smaller differences between substances can be detected than would be possible if only a few steps occur , as in Wilcke 's list of nine materials .
In the present section we need only notice column 2 , which shows the materials used in normal order .
These are in all cases the ordinary commercial article , but , in order to establish standard materials for future reference , particulars will be given in some cases .
Furs.\#151 ; In rabbit 's fur the hairs are clipped short , in imitation of ermine .
In oppossum 's and cat 's fur the hairs are of natural length .
These furs have been selected from a great number , as giving convenient steps between places 1 and 12 in the Table .
Vitreous Silica , Quartz , and the Glasses.\#151 ; In places 3 to 9 are shown no less than seven specimens of this family .
The quartz is the natural material to which has been given an artificially polished surface .
The vitreous silica used has surface smooth from fusion .
Three glass surfaces , including glazed porcelain , are smooth from fusion ; one is artificially polished , and one surface ( place 9 ) is the edge of thick plate glass , obtained after the plate has been cut by diamond and broken in the usual way .
Wool , Silk , Cotton.\#151 ; Care must be taken in using fabrics purporting to be made of these materials , especially in the case of silk , which is so often adulterated with other yarn .
The standard materials in which I trusted are skeins of yarn .
The silk and wool came from yarn merchants , guaranteed pure .
The silk has the natural gum washed quite free from it .
The cotton used is pure white cotton-wool .
Mica , Calcite.\#151 ; The natural surfaces of fresh specimens are used after recent cleavage .
The surfaces of salt crystals are scraped before use .
Metals and Alloys.\#151 ; Twenty-three materials are quoted .
Their surfaces are used scraped or rubbed with sand-paper , or , where possible , burnished .
It is important to avoid surface oxide , sulphide , etc. , but no differences have been detected in any metal surface on account of the above different methods of cleaning .
Felt ( mixture of hair and wool and other fibres).\#151 ; The specimens are of c 2 20 Dr. P. E. Shaw .
close texture , about 5 mm. thick .
The material is such as is used for carpeting .
Vulcanised Fibre.\#151 ; This red composite material is supplied rolled in sheets and tubes .
A tube is polished by rotation in the lathe under the pressure of another tube of the material .
No liquid or other foreign material is used in the process .
Woods.\#151 ; Ebony is the only wood of many tested , which does not occupy place 19 .
The specimens are planed up smooth and parallel to the grain .
Shellac , , Sealing Wax , Resin , Sulphur , are melted and allowed to solidify on glass tubes .
The two former must not be raised far above the melting point in the process .
This care is necessary to avoid disintegration .
Ebonite.\#151 ; The solid rod is polished in the same way as vulcanised fibre .
Carbon.\#151 ; Gas carbon rods , as used for arc lights , and charcoal of various kinds have all the same place ( 29 ) in the Table .
Celluloid.\#151 ; This also is best in its purest form in limpid sheet highly polished .
The opaque , less pure , form is higher , about place 29 . .
Indiarubber.\#151 ; Under this head is implied the pure , translucent , very elastic , sheet rubber used for mending the inner tube of bicycles .
There are many forms of impure rubber ; for instance , white rubber is much higher in the Table ( place 29 ) .
It is noteworthy that differences of behaviour are easily observable between insulating materials all of one class .
In order to indicate how differences in the surfaces not detectable by any other means are easily shown by tribo-electric effects , take a few cases :\#151 ; ( a ) A highly insulating material , such as celluloid or mica sheet , being cut into two pieces , it will generally be found that one piece will be + and the other \#151 ; when rubbed together .
There is nothing whatever in their appearance to show that these pieces cut from the same sheet are in any way different .
( b ) Nine different furs tested were found to take definite positions between places 1 and 13 in the Table , various hard siliceous materials falling into place between them as shown .
Furs .
Woods .
( Asbestos .
) Black wolf .
Babbit .
Caracule .
Fox .
( Vitreous silica .
) ( G-lass .
) ( Glass .
) Opossum .
Astrachan .
Mink .
Cat 's skin .
Skunk .
( Matt mica .
) Beech .
Bed deal.- Jarra .
Canary .
Walnut .
Box .
Mahogany .
White deal .
Lignum vitae .
Bose wood .
( Matt glass .
) Oak .
Ebony .
i Experiments on Tribo-Electricity .
( c ) Of 12 different woods , all except one ( ebony ) form one group , place 19 , yet when rubbed together they can be arranged in definite sequence like furs .
The lists of furs and wood are given on p. 20 .
III.\#151 ; Apparatus and Methods .
The confusion which has been a feature of this subject in the past may be attributed to the uncertainty as to materials used , their past treatment , and the methods employed in experimenting on them .
With the attempt to standardise the methods , or at least to specify my own processes , I will give particulars throughout , even though these are admittedly simple .
The materials used may be divided into conductors and insulators .
The latter can , in general , be used unmounted ; the former ( including partial conductors , e.g. , wood , cotton fabric , asbestos ) must be mounted on a nonconductor .
In general , those rods , tubes , or sheets , which have to be mounted are attached by sealing-wax to the inside or outside of glass tubes .
It is necessary to have sufficient distance , say 10 cm .
, between the hand and any conductor undergoing test , for if the hand be nearer it may remove , by induction , any free charge on the material .
Easily fused materials , e.g. .
shellac and paraffin wax , are mounted in a solid lump at the end of a glass tube .
Lumps of material , such as minerals or carbon , are held in a clip specially made of ebonite .
Mercury is used in a cylinder into which any material can be plunged .
The gold-leaf electroscope used is provided with a shallow horizontal cylinder open at both ends .
For charging the electroscope a wire is arranged close to the cylinder .
This wire is connected with a live wire of an electric light circuit at 205 volts .
As a means of discharging the surfaces , a small bunsen flame is kept alight in front of the observer .
The flame itself should never be used as a discharging agent , since most surfaces become abnormal very quickly in the flame .
Round the flame is a zone , the discharge zone of air , so highly ionised that any charged body moved about at , say , 10 cm .
from the flame for a few seconds becomes completely discharged .
The gases rising from the flame are more highly ionised than the air at its side , but there is great danger , if the body be plunged in these gases , that the temperature of the body will rise so far that it will become abnormal .
In investigating critical temperatures ( column 4 in Table , p. 32 ) , the substance is heated by being placed on the upper shelf of a copper oven , the oven being swaddled in sheet asbestos .
The door of the oven is closed and the temperature is raised by bunsen burners underneath , and is recorded by a thermometer inserted through the top of the oven .
It is easy to raise Dr. P. E. Shaw .
the temperature slowly to , say , 320 ' C. , which is high enough for present purposes . .
As to the condition of the surfaces used , cleanliness is desirable , though apart from chemical impurity , e.g. , oxides , sulphides , on metals , dirt does not influence results as much as might be supposed .
It is clear that if , for instance , there is a layer of dust on an ebonite rod , this , when rubbed with silk or paper , will be brushed away at the first stroke , and the true surfaces will be brought into contact .
But metals must be used free of corrosion by the application of sand paper or by the process of filing or scraping .
Apart from metals , it is well to keep such materials as ebonite , gutta-percha , and vulcanised fibre clean by frequent scraping .
It is important to observe that materials , e.g. , mica , glass , or celluloid , supposed to be polished are really used in this state .
Polished celluloid and ebonite are easily rendered matt at the place of rubbing when pressed on some hard surfaces .
We must observe a uniform method of rubbing the surfaces .
Consider , as an example , the case of quartz-silk .
If the silk be brushed over the quartz it becomes + , but if rubbed hard on it its charge becomes \#151 ; .
These effects will generally be found when hard surfaces are rubbed with flexible ones , such as fabrics and furs .
To take another instance ; if a mica sheet be beaten on a glass tube it becomes slightly + , but if rubbed on it strongly \#151 ; .
The latter is the standard way , which I always use , hence in the above examples we place quartz above silk and mica below glass in the Table .
Purse themselves are somewhat uncertain .
The same specimen of opossum fur may at one end be + to vitreous silica , and at the other be \#151 ; to it .
In such cases as this the experimenter must learn the peculiarities , if any , of the materials he uses .
As will appear later the surfaces of most materials are susceptible of several states , all more or less stable .
Tims , if a smooth glass rod be under test it may have its position in the Table at place 3 , 5 , 8 , or 26 .
But repeated rubbing , gentle heating , and the slow changes which occur with lapse of time will bring the glass to a definite position , say place 5 , which is taken to be the position of glass in its most stable state .
It might be thought desirable to use special devices to clear of adsorbed layers the surfaces about to be rubbed .
For instance , well-known methods could be suggested , such as ( 1 ) tearing off the surface by gelatine films , or ( 2 ) scraping in vacuo .
These processes have been found necessary in other surface experiments , such as surface tension and photo-electricity .
But there is an essential diffei'ence between the delicate forces at work in the latter subjects and the rough impact of solid on solid , such as occurs in Experiments on Tribo-Electricity .
tribo-electric experiments .
In this case , the true surfaces would bear on one another .
The actual bearing surface of one solid on another would be so small that the pressure per square millimetre would be enormous , and would suffice to break through any thin outer foreign film , especially one of a liquid condition .
IV.\#151 ; The Full Tribo-electric Series .
At the end of this paper is a Table which gives in compact form the experimental results of this investigation .
In the second column are shown the materials used , arranged in 36 standard places .
Those in capital type I have found more useful than the rest .
The metals are italicised .
The whole list from Asbestos at the + end to India-rubber at the \#151 ; end may be likened to a staircase of 36 steps , which steps are , of course , unequal in depth to an unknown degree .
The list itself could be expanded apparently without limit .
Group B follows Group A without break . .
Column 3 contains such materials , bright when in column 2 , as can be rendered matt by some abrasive , e.g. sand paper .
It will be observed what a vast disparity there is , in some cases , between the surfaces of the same material when polished and when matt .
Thus , the members of the glass family , normally round about place 5 , are , when matt , at place 20 .
In column 4 are the places of the materials when these have had their surfaces raised above the critical temperature , which temperature where it has been found is inserted in brackets .
I call a surface abnormal when , as a result of being heated above a certain temperature , it moves from its normal place in column 2 to some other place in column 4 .
This particular temperature I call the Critical Temperature ( C.T. ) .
There are two ways of making a surface abnormal : ( 1 ) By raising the substance as a whole to the C.T. in the air-bath ; ( 2 ) by passing the surface rapidly through the hot body of a flame .
In the latter case ( which , of course , is not applicable to all materials ) , though the interior of the substance may remain cool , the surface particles will be raised to the C.T. in a few seconds or less .
Experience shows that the abnormal place of a body in column 4 is always the same by whichever of the above processes this condition is obtained .
Suppose a body has been raised to the C.T. , and on being rubbed at that temperature is found to be abnormal at place 26 .
It will remain abnormal when cold and still occupy place 26 in the Table .
It will be observed that the critical temperatures range from 300 ' for quartz ( a hard substance ) to 70 ' for vulcanised fibre ( a soft compost ) .
In general , the minerals have high critical temperatures , while in metals the C.T. ranges from 240 ' to 130 ' .
As to the accuracy of these figures , they are Dr. P. E. Shaw .
trustworthy to 5 ' in the case of glass and vitreous silica , which have received special attention .
For the other materials , the figures are probably reliable only* to 10 ' or 15 ' .
There is some doubt about the case of tin .
It melts at 230 ' , and when raised to 210 ' it certainly took the place shown in column 4 , but as at that temperature the surface was yellow with oxide , this particular result must be considered doubtful .
Column 5 has only six entries .
Many more have been observed , but as they are not quite consistent , they are omitted .
Suppose a sheet of cork or ebonite be pressed and rubbed by an ordinary " flat-iron " when at a temperature , say , about 150 ' , it will descend in position as shown in the Table .
Another method is , after warming a sheet , to place it between two sheets of its own material , and press the whole together while the hot sheet is pulled from between the others .
This is called the Pressed-Hot ( P.H. ) state .
It can be removed , in the case of hard bodies , such as glass , by passing the surface dexterously through the flame for a short time .
If the surface be heated too long , it will go further and become abnormal .
Columns 6 and 7 should be considered together .
In experiments to be described later ( p. 26 ) , it is shown that , under certain conditions of stress , a convex surface and a concave surface , both of one material , become charged in opposite senses .
It will be seen that mica behaves in one way ; paper , ebonite , and celluloid in the contrary way .
Thus the Table ( p. 32 ) shows how readily the surfaces of a great variety of materials change their places in the series .
A scrutiny will reveal that these changes occur according to very simple laws .
Divide the second column in two groups , A , B , by drawing a demarcation line below place 14 .
Then it is observed that materials in Group A all move one way ; those in Group B all move the contrary way .
The laws of these movements may be tabulated thus:\#151 ; Matt .
Abnormal .
P.H. Convex .
Concave .
Group A Down Down Up Down Up .
Group B Up Up Down Up Down .
Another peculiarity in this grouping is that all the animal materials occur in Group A ; all the vegetable materials in Group B. The many empty spaces occurring in columns 3-7 are not blank through inadvertence .
Practically all the suitable materials have been tested for these effects , but it has not been possible to perceive effects for all the diverse substances in the Table .
In no case has any distinct result been !
Experiments on Tribo-Electricity .
suppressed .
But there have been a few cases where effects have been detected , but have been subsequently ignored , as they would not repeat .
Thus glass , silica , and quartz have sometimes been obtained P.H. at place 2\#151 ; but this effect is not consistent .
The contrary influences found in the two groups I call duality .
Notes on the Behaviour of Mercury .
Before concluding the description of the Table ( p. 32 ) , mention should be made of the anomalous behaviour of mercury .
It will be seen that this metal is placed with platinum and gold ( place 34 ) , and also high up ( place 2 ) .
The effects observed are as follows:\#151 ; ( 1 ) When rods of various materials are gently pushed into mercury , every one in the series , except celluloid and indiarubber , will become charged + .
Thus we assign to mercury place 34 .
If , however , the rods be plunged quickly into the mercury or stirred in it , they all , except asbestos , become \#151 ; .
Let us denote mercury when the friction is violent as ( mercury ' ) .
Its place is 2 .
Possibly violence causes the rods to become extremely abnormal in contact with mercury .
Lenard 's theory ( 9 ) regarding " waterfall " electricity may be recalled .
In the case of water , he supposes there is an electrified double layer , the external one being \#151 ; and the internal + .
If there be a similar double layer in mercury , the external one would be + and the internal \#151 ; .
( 2 ) Abnormal glass , polished or matt , is \#151 ; to mercury even when immersed in it gently .
Yet the same specimens of glass are 4- to guttapercha , which itself is + to mercury .
The same effect will be found for abnormal mica and other materials .
Supposing the charges obtained when bodies are plunged into mercury are really due to friction , then the effect just recorded seems to mean that the abnormal surfaces are in reality below place 34 , for such gentle friction as mercury provides , but that rubbing with solid surfaces so flattens , or otherwise affects , the surface particles as to bring the abnormal surface up , in the case of glass , to place 26 .
( 3 ) A glass rod plunged into mercury becomes highly charged .
On repeating the process several times , however , the + charge becomes very small .
But the original excitability is restored by rubbing the glass by hand or by a cotton duster .
It is possible that the glass is rendered abnormal by mercury , and made normal again by rubbing .
( 4 ) If mercury be raised in temperature , its power of exciting charges on solids immersed in it gradually lessens as temperature rises , returning again entirely at normal temperature .
Thus , at 110 ' , the charge produced on ebonite is small , whether immersion be gentle or violent .
Glass in mercury at the above temperature receives no charge for gentle immersion , but Dr. P. E. Shaw .
receives a large charge when immersion is violent .
When the mercury is at 200 ' , it will give no charge to any body under any conditions tried . .
Notes of the Peculiar Behaviour of some Flexible Sheets .
( 1 ) Take two strips , cut from the same sheet of transparent , polished celluloid .
Place one sheet on the other , and draw them between thumb and first finger , bending the strips as they pass between the fingers .
Now place the strips , still in contact , for a second in the discharge zone of a flame to remove the \#151 ; charges on their outer surfaces due to friction of the material on the fingers .
On separating the strips , that one of the surfaces in contact which was convex when flexed is + , the other surface in contact , the concave one , being \#151 ; .
An effect of this sort was pointed out by Jamieson ( 7 ) .
Hence opposite charges arise from complex strains on the opposed surfaces .
There is no special significance in using the fingers , identical effects being found if the fingers are replaced by other rubbers , say pieces of wood .
The above effects are recorded in the Table ( p. 32 ) , columns 6 and 7 .
( 2 ) Similar effects are found for two sheets of ebonite or cork , of filter paper or mica .
But it is of interest to notice that mica , which is in Group A , acts in one way , whereas the other materials , which are in Group B , behave in the contrary way for convex and concave surfaces respectively .
( 3 ) A variant of the above experiment is curious .
Place the two celluloid strips in contact , and draw them between the fingers as before , but in this case leave them unflexed in the process .
After removing the external \#151 ; charges as before in the discharge zone , the strips will be found to cling together strongly , and , when tested , are found to be charged , one + , the other \#151 ; .
This , of course , is a simpler experiment than ( 2 ) , for now the two strips are treated identically throughout .
The two surfaces in contact must differ in composition or condition , so as to act on one another , in effect , as different materials .
It is a commonplace in tribo-electricity to find two surfaces apparently identical which , when rubbed together , excite one another with opposite charges .
The present effect seems to be merely an example of this kind .
In experiment ( 1 ) the convex surface becomes always + to the concave , however the surface of the two strips be changed about .
But in experiment ( 3 ) the effect is peculiar , not to the material , but to a given specimen : if the surface is + in one position , it will be -f- always , whatever its position relative to the other surface .
Thus , the effect in ( 1 ) depends on flexure , whereas in ( 3 ) it is simply characteristic of the materials used .
Experiments on Tribo-Electricity .
Y.\#151 ; Theory .
On the simple electronic theory of tribo-electric phenomena there are free electrons close to the surface of solids which readily leave the surface when rubbed .
The materials high up in the series are supposed to be characterised by having electrons specially free to leave .
Tins theory , as it stands , does not account for the presence in the Table , mixed in among insulating bodies which are poor in free electrons , of metals and alloys , all peculiar for their richness in free electrons .
Nor is it evident how it can explain the effects in columns 3-7 in the Table .
These are striking and consistent enough to call for an explanation .
A theory is wanted .
Moreover , a definite concept of surface conditions is needed for continued successful investigation of the subject .
The outstanding results are : ( 1 ) the order ( not unmethodical ) in which materials arrange themselves in column 2 ; ( 2 ) the duality found in columns 3-7 in the behaviour of the top and bottom sections of the series ; ( 3 ) the principle of the critical temperature .
In commencing an enquiry into the basis of these effects the question arises as to any possible relation of chemical composition of the material to its position in the series .
In the following list is given the composition of the materials used .
They are arranged roughly in the order of column 2 , except in those cases where like bodies are grouped together .
Material .
Composition .
Group A. Asbestos ... ... ... ... . .
Furs ... ... ... ... ... . .
Wool ... ... ... . .
Silk ... ... ... . .
Felt ... ... ... ... ... ... Hand .
Silica ... ... ... ... ... Glass ... ... ... ... ... .
Quartz ... ... ... ... ... Mica ... ... ... ... ... . .
Calcite ... ... ... ... . .
Borax ... ... ... ... ... .
Fluor spar ... ... ... ... Pb , Al , Zn , Cd , Mn , Cr .
Group B. V ulcanised fibre Filter paper . .
Cotton ... ... ... Woods ... ... ... Cork ... ... ... .
Magnalium _____ Magnesium silicate .
I Compounds of C , H , O , N , possibly P , S. The surface of \gt ; hairs consists of flattened dried cells , containing a dehydrated proteid , ceratin ( C27H55OH ) .
\gt ; Si02 , with ( in glasses ) sometimes !
Na20 , K20 , CaO or A1203 .
K , Al , SiOo .
CaC03 .
NaoB40,10Ho0 .
CaFo .
J*Cellulose ( C6H10O5 ) " with traces of mineral matter .
( 70 Al , 30 Mg ) .
Dr. P. E. Shaw .
Material .
Composition .
Group B\#151 ; ( contd .
) .
Para.ffin wax Hydrocarbons , e.g. C30H62 .
K2Cr207 .
1 \#166 ; Potassium bichromate ... Amber Shellac , [ Mixtures of compounds of C , H , 0 , of unknown formula .
[ Sealing wax sometimes has mineral matter in traces .
Sealing wax " Resin Potash alum KA1(S04)2,12 H20 .
KCr(S04)2,12H\gt ; 0 .
NH4Fe(S04)2,12H20 .
Aluminium silicate ( hydrated ) .
| Hydrocarbons and sulphur .
Similar to indiarubber , but without sulphur .
FeS2 .
Chrome alum Iron alum Slate Ebonite Indiarubber Gutta-percha Iron pyrites Galena pbs !
C. Cu , Sn , Sb , Co , Ni , Bi , Pd , Ag , As , Te , Pt , Hg , Au Celluloid \#171 ; ... Nitro-cellulose ( C , II , N , 0 ) and camphor .
It is easy to detect characteristics special to each group .
Group A contains :\#151 ; \#166 ; ( 1 ) Silicates , including the glass family , asbestos and mica .
( 2 ) All the animal materials .
, ( 3 ) The notably + metals and lead .
Group B contains:\#151 ; ( 1 ) No silica compounds , but a great variety of carbon compounds .
( 2 ) All the vegetable materials .
( 3 ) Many metals , including those notably \#151 ; .
There is no apparent gradation in density , but the hard silica compounds in Group A are in striking contrast to the soft bodies , especially the waxes , in Group B. Hydrated bodies are found in the lower rather than the upper group .
The three alums are seen to be widely separated from one another in the order we should anticipate from the electro-potentials of their distinguishing metals .
Peculiar interest attaches to the elements , metallic ( 20 in number ) and non-metallic ( 3 in number ) , in the series .
Of the metals , all but two , lead and mercury , are arranged according to the electro-potential series of Ostwald( ?
A ) .
Allowance must be made for the fact that owing to their conductivity metals rubbed together show no charges detectable by the present means ; hence the metals are grouped in many cases as shown , no less than nine occurring at place 29 .
Mercury with electro-potential \#151 ; 1*03 is so i Experiments on Tribo-Electricity .
nearly the same as silver ( \#151 ; 1-05 ) and palladium ( \#151 ; 1*07 ) that its displacement in the series relative to these two metals may be ignored .
This leaves lead as the only exception to the electro-potential order .
Lead is undoubtedly peculiar , for with potential \#151 ; 013 it is above , for instance , aluminium with potential +1*00 .
Three distinct specimens of commercially pure lead behaved alike ; and no change in behaviour was found after the metal had , been melted and when cool cleaned on the surface .
Commercial sheet lead is so pure that it is not considered necessary to try a specimen of special purity .
An independent reason for supposing lead to be in Group A is that when abnormal it descends in the series to about the position of undoubted Group A metals , e.g. aluminium and zinc .
Apart from the metals , only three elementary solids , carbon , sulphur , and tellurium , have been tried .
These highly negative elements occur , as we should expect , near the bottom of the series .
One supposes that the other non-metallic elements , if obtainable in suitable solid form , would also take their place at or beyond the bottom of the list , and that the most strongly electro-positive elements , potassium and sodium , would occur high up in the series .
In general the basigenic elements come above those with oxygenic qualities ; yet the property of forming very stable oxides , sulphides , etc. , would seem to be a better criterion of tribo-electric position than base- or acid-forming power .
Thus if calcium or manganese be rubbed with their natural oxide on the surface they will take places far down in Group B ; if cleaned , and the pure metal exposed , they are found in Group A. Again , silica is a prominent compound at the top of the series , whereas the inert metals , platinum and gold , are at the bottom .
Now , if acid-forming power were considered , silicon should be low in the series and silica also very low , certainly not far above platinum and gold .
These matters appear to be vitally important from the theoretical standpoint .
As to the critical temperature , the change from the normal to the abnormal state appears to *be sudden .
A rise of 10 ' or so suffices to take the surface from one state to the other .
But glass at 220 ' ( i.e. 23 ' below O.T. ) has been found at place 20 in the series , as if it were in a transition state from normal to abnormal .
In the cases of some metals , e.g. , silver and copper , the abnormal state has been looked for , but not detected .
It is clear that in most metals it would be hopeless to look for a critical temperature higher than , say 300 ' , as oxide would then form readily on the surface .
Closer and more extensive research may reveal the significance of these critical temperatures .
In scrutinising the series , one can detect nothing in the ordinary physical or chemical characteristics to elucidate the phenomenon of duality mentioned on p. 25 .
Dr. P. E. Shaw .
The conditions at the actual surface of a solid are , of course , peculiar .
The particles there are more free than below the surface .
They are , in general , unprotected and , so to speak , raw and liable to strain , orientation , and disintegration by outside influences such as the impact of other solids .
Tribo-electric charges are derived from the actual surface ; but whether from constituent parts of the atom or from interatomic or intermolecular sources is not decided .
The theory will be very complicated if , as is possible , the charges are derived from two or more of these sources .
The atom is the simplest unit , and would appear to be the most likely source , but the question * must be left open at present .
Now , if two groups ( e.g. , Groups A and B in the Table ) of bodies behave in ways exactly contrary to one another , one might imagine each group to be characterised by its own kind of unit ( whether atomic or molecular ) .
Suppose the characteristic of Group A is an a unit , that of Group B a ft unit .
If we assign certain qualities to the units , then in any surface the units would decide how the surface would change its place in the series when its condition is changed .
According as the a or ft units predominate in the material the action will be that of Group A or Group B in the Table on p. 32 .
The case has been stated here in the most general way .
It would be premature to go further and outline the model which I have had in mind throughout the experiments , and which has successfully indicated the course to pursue in these investigations .
Something should be said about the principle of the electric double layer by which many physicists , e.g. J. J. Thomson ( 8 ) and P. Lenard ( 9 ) , have interpreted surface effects .
Amongst others , 0 .
W. Bichardson ( 10 ) has explained the fact that in thermionic experiments a definite amount of work must be spent to enable an electron to escape from a heated solid surface .
The view is generally held that the \#151 ; layer is always outermost of the two layers .
Again , R. A. Millikan ( 11 ) has given new support to the principle by proving the exactness of Einstein 's formula \my* = h(v \#151 ; v0 ) for photoelectric effects on certain metals .
Recently , J. Frenkel ( 12 ) has developed the idea .
Assuming that all positions and all orientations of an atom at the surface are equally probable , he explains the double-layer principle on the simplest basis .
For if Rutherford 's nuclear atoms be placed in all positions and all orientations at the surface , the electron orbits will in the aggregate project beyond the + nuclei , and there will be established a \#151 ; layer outside a + layer .
Frenkel carries on the argument , successfully applying the principle to a great variety of surface effects .
In short , the evidence in favour of the double-layer effect seems overwhelming .
i Experiments on Tribo-Electricity .
However , the results of the present paper may be taken to show that whereas all orientations at the surface are equally probable for the abnormal state , and less so for the matt state , yet for the normal and pressed states the surface particles are orientated unequally ; a ' quasi-mechanical strain being imposed ( by rubbing ) on the surface particles .
As an analogy , imagine a row of pivoted magnetic needles all pointing N. , but their centres in line E. and W. If the pole of a bar magnet be taken along the line near but not touching the needles , these will , if at suitable distances from one another , be left at the end of the operation pointing in line E. and W. The bar magnet has orientated them , and they are left in a stable condition of strain .
Jolting the system would cause a release from the strained state of the needles , just as heat causes the release from the normal to the abnormal state in the case of the tribo-electric experiments .
Orientation of surface atoms would be tantamount to bringing the outer \#151 ; layer and the inner -f layer nearer to one another .
Another type of atom , though not apparently one of the Rutherford pattern , might produce at the surface of a solid a double layer with the + layer outermost , If this were so , we could explain the effects of Group A , by imagining the new type of atom predominant in those materials ; and the effects of Group B by having the Rutherford atom predominant in the materials of this group .
Whether , in the double layer , the \#151 ; layer or the + layer be outermost at the surface , the layers would be most separated from one another in the abnormal state and least separated in the pressed state .
VI.\#151 ; Conclusion .
The endeavour is made to direct attention to a long-neglected field of research .
The results revealed are considered sufficiently novel for publication as they stand .
With more refined experimental methods the subject will advance no doubt from its present stage ; but the process will be slow .
Quantitative work on materials in vacuo cannot easily be done when , as here , considerable mechanical force has often to be applied to the materials .
But this is not the only reason for expecting slow progress .
The materials possible to be tested are unlimited , and generalisations should not be based on a few instances only. .
In the last few years the surface conditions of solids have been investigated from new standpoints .
To mention two instances bearing on our experiments : H. L. Curtis ( 13 ) has observed enormous change ( as much as one hundred million-fold ) in the surface conductivity of insulators as the relative humidity of the air over the surfaces varies from 20 to 90 per cent. Dr. P. E. Shaw .
This brings in the moot subject of adsorption , which one would expect to have a bearing on tribo-electric effects .
Mr. C. Hayes and I have observed a direct connection between dampness of a surface and the sign of the charges it gives when rubbed .
Again , J. W. French ( 14 ) has discovered that the surface of artificially polished glass to a depth of 5 ya is of a peculiar nature , differing in cohesion from the main underlying material .
Mr. Hayes and I have found that the surface of polished glass has under some conditions a neutral state .
In this state it yields neither + nor \#151 ; charges .
This effect is found only in the artificially polished material .
Any theory on this subject to be acceptable should embrace not only the effects included in the Table , p. 32 , but also several others such as the varied results obtained ( see p. 22 ) according as one surface is brushed , rubbed , or struck by another .
The theory should take cognisance , also , of the peculiar behaviour of mercury ( p. 25 ) , though as mercury is the only liquid used in these experiments its case is singular .
This paper is meant to be the first of a succession on the subject of tribo-electricity .
Some quantitative work has already been done on various branches of the subject , e.g. , the influence on the surfaces about to be rubbed of ( 1 ) low vacua ; ( 2 ) electric high-potential discharge ; ( 3 ) various liquids .
Mr. C. Hayes , M.Sc .
, has worked with me for a considerable time on these subjects .
I am glad to acknowledge valuable help in obtaining data for the present paper from him and from Miss H. M. Browning , B.Sc. I also wish to acknowledge my indebtedness to Prof. G. A. Schott , I).Sc .
, for valued advice as to theory .
VII.\#151 ; Summary .
( 1 ) The tribo-electric series , in which solid materials are arranged in order according to the charge they acquire when rubbed together , is reliable with due precautions .
In the Table are shown the common solid metallic and non-metallic elements , interspersed with a great variety of compound materials .
( 2 ) Most solids are found to alter their place in the series if heated above a certain temperature , , which is specific for each material .
This temperature is called the critical temperature .
The surface in its new condition is termed abnormal .
( 3 ) The series may be divided into an upper .Group A and a lower Group B. It is found that these groups have tendencies contrary to one another as the surfaces of the materials are rendered { a ) matt , or ( b ) abnormal , or ( c ) pressed , or ( d ) flexed .
If under any of these agencies Group A becomes more +-forming , Group B becomes more --forming , and vice versed .
This principle is here called duality .
It is possible that the Dr. P. E. Shaw .
[ .To face p. 32 .
A Table showing the Tribo-electrio Normal Series , Column 2 , with Variations for Matt , Abnormal , Pressed and Flexed States in Columns 3 , 4 , 5 , 6 , and 7 respectively .
The more important materials are in capitals .
Metals are italics .
Group A. Group B. Place .
Materials in normal Pre in sir mred att Paces .
Abnormal ( and critical P.H. Flexed .
Place .
Materials in normal Prepared Abnormal ( and critical temperature in ' C. ) .
P.H. Flexed .
series .
temperature in ' C. ) .
( P- 24 ) .
Convex .
Concave .
series .
matt surfaces .
( p. 24 ) .
Convex .
Concave .
+ 15 Filter Paper .
1 Asbestos ( sheet ) .
16 Vulcanised Fibre V. fibre Ca ( 240 ) ?
F. paper .
17 Cotton Caleite(270 ) .
2 Rabbit 's Fur , Hair 18 Magnalium Mica ( Hg ) ' .
19 Potash alum , rock salt , Woods , V. Silica satin spar 3 G-lass ( Combn.-Tube ) .
20 21 Woods , Fe Unglazed porcelain , sal- Quartz Glass V. fibre ( 70 ) F. paper .
Woods .
4 Vitreous Silica , Opos- 22 ammoniac Potassium bichromate , Combn .
Tube Ebonite(lOO ) Pb ( 240 ) .
sum s Fur .
Paraffin , tinned iron .
Glass ( fusn .
) 23 Cork , ebony Al ( 240 ) Cork .
5 Mica Mica .
Zn ( 220 ) .
24 Amber Fe ( 200 ) .
Cork .
6 7 Mica .
Wool .
25 Slate , chrome alum Sealing Wax .
Resin Shellac .
Cork .
Mica .
26 Shellac , Resin , Seal- V. silica 245 .
8 Glass ( pol .
) , Quartz * ing Wax Glass ( 245 ) .
Mica ( 270 ) .
Quartz ( 300 ) .
( pol .
) , glazed porcelain .
* Brass .
9 Glass ( brn.-edge ) , ivory .
27 Ebonite Ni ( 200 ) Ebonite .
10 Calcite Wool Zn . .
28 Ebonite Sb ( 130 ) .
Sn ( 210 ) ?
11 Cat 's Fur .
29 Co , Ni , Sn , Cu , As , Pi , Ebonite .
Ebonite .
Sb , Ag , Pd , Carbon , Te , eureka straw , Ca , Mg , Pb , fluor spar , 12 30 copper sulphate , brass Para Rubber , iron alum Celluloid .
Au ( 130 ) .
borax .
Silk .
13 31 Gutta-percha Pt ( 130 ) .
32 Celluloid Indiarubber .
14 .
Al , Mn , Zn , Cd , Cr , Felt , hand , wash leather 0 1 FE Silk .
Cd ( 200 ) F. paper .
33 34 35 36 Sulphur .
Ft , Mg , Au Celluloid .
Indiarubber Sulphur Celluloid Celluloid .
Celluloid .
\#151 ; Experiments on Tribo-Electricity .
phenomena observed by Hesehous with radium may be another case of this duality of action , for glass is in Group A and ebonite in Group B. ( 4 ) Anomalous effects are observed when liquid mercury is used as one of the materials , its behaviour being quite unlike that of solid surfaces .
( 5 ) As to theory , it is suggested that the prevalent idea that the electric double layer , existing at the surface of solids , has the \#151 ; layer outermost in all cases is incorrect .
Normally , the materials in Group A would have \#151 ; outermost , those in Group B having + outermost .
Orientation of surface atoms would give rise to changes in the disposition of the two electric layers , and so account for observed effects .
( 6 ) Tribo-electricity undoubtedly affords a means , of extraordinary delicacy , of discriminating between materials apparently alike .
LIST OF REFERENCES .
1 .
Chwolson , O. D. , 'Traite de Physique , ' vol. 4 , Fasc .
1 , pp. 264-9 .
2 .
Wiedemann , G. , 'Die Lehre von der Electricitat , ' pp. 208-220 .
3 .
H6sehous , N. A. , 'Jurn .
Russk .
Fisik-Chimicesk .
, ' Obscestva 34 , pp. 1-15 ( 1902 ) ; 35 , pp. 575-80 ( 1903 ) ; 37 , pp. 29-33 ( 1905 ) .
4 .
Morris-Owen , ' Phil. Mag. , ' April , 1909 .
5 .
Morris-Jones , W. , 'Phil .
Mag. , ' February , 1915 .
6 .
Shaw , P. E. , ' Proc. Phys. Soc. , ' vol. 27 , April , 1915 .
7 .
Jamieson , W. , ' Nature , ' vol. 83 , p. 189 , April 14 ( 1910 ) .
7a .
Ostwald , W. , ' Lehrbuch der allegmeinen Chemie , ' vol. 2 ; also ' General Inorganic # Chemistry , ' by Alexr .
Smith ( Geo. Bell ) , p. 670 .
8 .
Thomson , J. J. , ' Camb .
Phil. Soc. Proc. , ' vol. 14 , p. 105 ( March , 1907 ) .
9 .
Lenard , P. , 'Ann .
d. Physik , ' vol. 47 , 4 , pp. 463-524 ( July 8 , 1915 ) .
10 .
Richardson , O. W. , ' Emission of Electricity from Hot Bodies .
' 11 .
Millikan , R. A. , 'Phys .
Rev. , ' vol. 6 , p. 55 ( July , 1915 ) .
12 .
Frenkel , J. , ' Phil. Mag. , ' April , 1917 .
' .
13 .
Curtis , H. L. , ' Scientific Papers , Bureau of Standards ( U.S.A. ) , ' No. 234 , June , 1914 .
14 .
French , J. W. , 'Trans .
Optical Soc. , ' November , 1916 , and February , 1917 .
VOL. XC1V.\#151 ; A. D

