]\gt ; The in HoldincJ Varyi'\yacute ; g of Part I by Professor F. T. , F.B.S. ; Part II by Professor 1 T. OUTON , , and Iiss B. POOL .
( Received November 30 , 1905 , \mdash ; Read January 25 , 1906 .
) .
The knowledge of the quantity of water held under varying circumstances by substances of an absorbent character , such as cotton or woollen material in an atmosphere of any given humidity , is not only of importance in rometry , but is also of general interest in connection with the processes used in drying such materials .
No investigations , however , of this subject seem up to the present to have been ever published .
Some years } , while making a series of comparative determinations of the weight of moisture absorbed out of the atmosphere by different kinds of fabrics , two points were noticed .
The first was that the weight of water absorbed or held by a given material under different conditions of moisture and temperature of the , appeared to depend only on the hygrometric state ( i.e. , the ratio of actual vapour pressure to the maximum possible ) , though of course the actual amount of moisture present in the atmosphere for the same ratio is very different temperatures .
The second point noticed was that as the atmosphere varied from saturation , the temperature remaining the same , the amount of water held followed some law much greater reduction in weight a given change in vapour pressure when near saturation than is subsequently obtained .
As described below , this law ultimately proved to be a simple parabolic one , at least until desiccatiou .
Earl . ?
.\mdash ; The first experiments were made in the way , they were subsequently found not to be capable of sufficient inement , as far as they they served to the above relationls .
The used was first desiccated in the presence of air over phosphorus pentoxide , and veight ascertained .
latter WitS done in the open air , so that mented through condensation of nloisture .
To the initial ( curve was ) lotted of time , and extrapolated the first moment of to the atnnosphere .
The material was simply left exposed ordinary atmospberic conditions , and its weight observed time to time , the hygrometric state of the atmosphere being found simulPressure in , etc. taneously by a wet and dry bulb instrument .
These determinations were carried on for some months , a reading being made each day .
From these a few of the more cases are iven in Table I , illustrating the first relation .
These have been selected in pairs , so that the weight ( W ) held in the material is approximately the same , while the amount of moisture in the air is very different , as shown by the vapour pressu1e ( p ) , but in each case it will be seen that the hygrometric ratio is also approximately the same , i.e. , the ratio of ( P ) the vapour pressure in the atmosphere to the saturation pressure for the temperature ( t ) at the time .
This relation may then be shortly written as .
Table I. Subsequently }ccurately devised experiments , carried out by Miss B. Pool , and given in Part II of this paper , have proved this relation , which indeed can be shown from thermo-dynamic considerations to necessarily hold , at least within usual atmospheric ranges .
Thermo-Dynamic Considerations.\mdash ; To show this , some of the material may be imagined placed under a piston in a cylinder water vapour , and a cycle of the usual type passed through .
Let the material holding the requisite weight of water be in equilibrium with the vapour a pressure less than the maximum of saturation pressure .
First an isothermal expansion takes place from volume to , evaporation of fresh vapour from that in the material will take , but if the amount required is small in comparison to that in the nlaterial , the pressure may be taken as constant .
Let be the weight thus evaporated .
The heat required for the evaporation is , where is the latent heat of vaporisation under the conditions of the operation .
Secondly , let an adiabatic expansion cool the contents of the cylinder S degrees , the pressure falling in consequence .
Thirdly , 294 Prof Troutoll and Miss Pool .
Pressure [ Nov. 30 , the is isothermally compressed the proper amount to admit of the cycle completed in the fourth by an adial ) atic compression .
We thus have , Carnot 's principle , But where is the of the vapour at temperature .
Hence is same in the two cases , where is the saturation pressure at temperature .
so we may write That is to say , the slol ) of the givin for different temperatures the pressure in equilibrium with the nnterial wetted with a constant weight of water , is to that of the rdinary vapour curve in the ratio of the ordinates which represent the pressures .
This necessitates the pressure to the saturation pressure , or .
That is to say , however the telnperature , if the hygrometric state is kept constant , the tht of water held by the materia ] remains constant .
It is convenient to have a name curves on the pressure-temperature diagram , which are drawn so that the pressure is always a constant fraction of the sattllation pressure .
These curves throughout this paper will be called sohygrom i curves .
It is also convenient to have a term for the curve drawn giving at different temperatures the pressure of vapour from a stance wetted with or a constant weight of water .
These curves are called in the paper isonerv curves ( from moisture , lreftdy in use in aneroid ) , or shortly isoneres .
The considerations given above would show that approximately the isoneres coincide with the rometrics .
It will be shown later that in the case of solutions , when one loody is non-volatile , the same relation may be considered approximately to hold .
Second \mdash ; Some indication of the second relation , namely , the function is of , was got from the original set of experiments , although the range under atmospheric conditions is ssarily very limited .
As typical of the general results , the II is given , which is compiled from observations made with a specimcn of flannel .
Owing to the uncertainty of the ) servations , and Cion a themselves ( most probably due * See fig. 9 .
1905 .
] in with Holding Moistn , re .
to the unsatisfactory method adopted for ascertaining the vapour pressure , namely , the wet and dry bulb thermometers ) , they have been lnassed in groups round a mean value of the corresponding hygrometric states .
In the first column is iven the mean hygrometric state of each group ; in the second column tlJe corresponding mean weight of the flallnel ; while in the third column are the calculated values of the square of the diffel.ence in weight between the latter and that of the flannel ot saturation divided by Che difference between 100 and the mean rometric state .
These are seen to be approximately constant , except in the first two or three cases .
This discrepancy may be due to there being insufficient time for equilibrium to set in , because from subsequent experiments it appears that the discrepancy from the parabolic law tends rather to the other side at low pressures:\mdash ; able I H. S. W. H. S. pparntus U method described dependent upon the in the atmosphere , is obviously inconvenient and uncertain , as among othel sources of error , during rapid atmospheric conditions there is not time for equilibrium to set in .
In order to make experiments under conditions of complete control in respect to the amount of nloisture present in the surrounding atmosphere it is necessary to ' in an enclosed space .
renders it impossible to use the wet and dry bulb rometer for dete1mining the vapour pressure .
A form of dew point instrument was therefore used , in which the depositiou of moistule was ascertained by electrical means .
This form was adopted because the conditions of { { pparatus rendered it impossible observe by eye the moment of deposition in the usual manner .
The arrangement consists essentially of two endent 1ibbons or wires ( , platinum wound at a small distance apal t over sulface on which densation of the ) tnre is to take place , ) these are fused on the 296 Prof Trouton and Miss Pool .
Vapour Pressure [ Nov. 30 , glass .
The outside of a test tube is used for this purpose and a of iced water , maintained in the test tube by tubes leading from a served , when required , to produce the necessary in temperature for making an observation .
A sensitive galyanometer and cell are connected up with the ribbons as electrodes or poles .
While dry no current is indicated , but soon as deposition of moisture takes place one is observed .
It is of interest to note that previous to the regular observational " " deposition premonitory symptoms are always to be seen in the form of slight indications given by the galvanometer .
The vessel in which the material under exanlination was placed was of tin plate , double walled , and had a circulation of water constantly maintained through the space between the walls from a tank or kept at a constant temperature .
The temperature of the tank was kept constant by means of a thermostat placed in the vessel itself , and which controlled ) flame the boiler .
A small hole in the top of the vessel allowed of the nlaterial under examination by a fine wire from a balance , placed for the purpose overhead .
In this way the weight of water at any given of moisture of the atmosphere in the vessel could be determined .
A small aperture was also provided for introducing water when it was desired to alter the hygrometric state in the yessel .
Us various uses no very satisfactory was obtained with this arrangement , and it was subsequently abandoned an ratus in which , instead of the amount of water in the material for c , oiven hygrometric conditions , the is gradually increased by definite incremenbs by of water to the material and the corresponding equilibrium pressure observed .
In these experiments , in order to avoid other sources of error , no attempt was made to keep the temperature constant by a thermostatic arrangement , as in the previous apparatus , but instead the observations were reduced to a common temperature by means of the first relation .
The method consists essentially in the dried material a vessel freed from air , and into which can be as equal of watel until saturation is produced .
ressul .
of the pour when equilibrium established is read after each feed , and the experimental relation is obtained , ( .
reduction to a elnl ) erature , giving the pressure in equilibriull with the material when ] varying amounts of water , that to , the on ssnre d Tlle may be described sbortly as barometer with an to hold th ) vetted 1 under mination , tube lvith a 1905 .
] Equilibrium with ) Holding oisture .
cock serving to troduce from time to time a supply of water of measured amount to the material .
A general view of the ) is given .
The bulb contains the material .
From it a tube passes down into a bath of mercury to form the pressure .
A tube , , containing phosphorus pentoxide , leads to the air pump .
The measured feed of water is introduced by the tube B. The amount supplied at each feed is that held in the fine tube lying between the two taps .
It is necessary that the water so intlo uced should be free from dissolved air .
This is clearly effected , since it is supplied from a reservoir formed by the space over the at the of ometer \ldquo ; tube , and into which air free water can be originally duced .
The method of a feed is to shut upper tap , open the lower tap , raise the mercury dish which is connected by a flexible tube .
In this way the fine tube is filled with the water .
The lower tap is then closed ( the upper opened .
The tube is connected to the bulb by a sealed joint , the left when the is removed the of introducing the specimen of material under examination .
At the lier s of an experiment the water in the .
tube on the ] ) tap passes over , and is absorbed cdmost with violence , but the material gets nearly saturated the process is slow , ( llld it becolnes necessary to drive over the vatel .
from the fille tube lie carefully warming the lower end of the fine tube with a small flame .
FIG. 1 .
In setting up the apparatus fi.eshly boiled water is drawlt over on top of the mercury the feed tube , closed .
On lowering the llercury dish so as to leave a space over water , .
is invariably to be given off .
This air is theu ejected the .
It is necessary to repeat this process number of times before all the .
is removed .
] reparing for all ] ) eriIncnt it was found to require several days to dry the lnaterial , and it quice free from air , which apparently continues for some time to be given off by it .
Prof. Trouton and Miss Pool .
Pressure [ Nov. 30 , here is no further object in rid of all air than to avoid a troubletemperature correction .
results of the series of experiments , lasted veeks , are exhibited in Table .
These wexe made with piece of flannel ) square hing 7rammes .
In the column is ( the crht of water the sl1ll } the feeds up tu the time in the flannel , in terms of that held by the c tube taken as unit .
A celtain amount of the water supplierl was 1e fill bulb with time , III .
2 .
0.098 0.150 1.287 1.669 1.721 14 .
1.721 I4 .
this .
small conparc ) the mount in .
pressure ined .
each feed .
At least 1 ] ) was made , so as to insure equilibrium the column is ] temperature , and in th fifth ) ated b the truth lclalion , the atio of , for different ] ) the ratio of the saturatiou prebsules , , the ebcnt in the lesults to with ) bolic law as 1900 with Substances Holding Moisture .
mentioned ve .
This will be seen 011 examining , where the parabola been drawn , and also the experimental points from Table III .
and are the and pressure at saturation .
The agreement is good until the pressure falls to about 20 per cent. of the maxirnum .
At this point there is a distinct indication of some further action or couling into operation .
Unfortunately complete had not been reached before beginning the experiments , so a second series was u1ldertaken to further examine this peculiarity of the curve at lower pressures .
The result of these experiments to confirm the accuracy of the first \mdash ; Isothermal at C. for Watel and series .
The action producing this deyiation from the }bolic 1 may perhnps be httingly referred to as the of a action , so that on adding water to the desiccated little or 110 pour ressure is Seri , the cond series the made complete .
The .
giveu in Table .
The first the same as ) .
the fifth colunm are given the pressures at Febrtl 17 , 1006 .
similar curve ) beeu obtained with cotton wool .
] 300 Prof Trouton and Miss Pool .
Pressure [ Nov. 30 , able I Date .
Weight .
1 9 ; 0.074 0.071 0.140 O. O. ; sure lcduced to C. 905 .
] in with Holding Date .
eight .
Feb. 28 15 16 .
1 19 2 .
20 28.9 Date .
eight .
Feb. 28 15 16 .
1 19 2 .
20 28.9 Date .
eight .
Feb. 28 15 16 .
1 19 2 .
20 28.9 Date .
eight .
Feb. 28 15 16 .
1 19 2 .
20 28.9 Date .
eight .
Feb. 28 15 16 .
1 19 2 .
20 28.9 0.669 0.704 0 .
0.998 1.0:31 24 0.992 1.196 26 .
as previousl by the firsC relation .
made in case , but in the points in fig. only lleall of three is used .
This series of observations does not lie ntinuous curve so satisfactorily as the first .
The first series 302 Prof outon and Miss Pool .
[ Nov. 30 , ths of Judy August , of the room in hich the servations were made did not ) , the second series was tnined druary { March , when ] artificial there wa , a in tempertltule .
small error in corresponds to a able v in ) pressure , and such an error is more to occur apidly f or atures .
at FIannel .
A fountl , ) for the ularitics , to fit ints unti ] the ssnrc fell to bout one-fifth then , until poinC , ) ) on th diagram .\mdash ; The is of interest ( to how ( is or naferial .
doubt held vities o the after csted 1 Kelvin 1905 .
] in Equilil)with Substances Holding Moisture .
many years , as well as in those spaces lying between and formed by two fibres .
The following considerations show , however , that these , all put together , can only account for a small portion of the moisture except the case of practical saturation , and that at lower pressures the moisture nrust be held by simple condensation on the surfaces of the material .
We can consider the cavities in the material , especially those between fibres , as equivalent on the to a certain conical , or rather cuspoidal cavity , which gradually fills up with water as the pressure of the vapour increases .
Though we Jnay not the shape of this cavity , we can , ving Lord Kelvin 's theory , calculate the diameter of the point which water reaches in it for each given pressure .
Thus where is the face tension at any selected teml ) ture , C. , the density of water , the maximum density of the vapoul , and the maximum pressure at the temperature If we knew the shape of the equivalent cuspoidal cavity we could arrive at the law connecting the weight held with the pressure .
for example that the cavity is conical we have for the law connecting and when is the angle of the cone .
This when plotted gives us a curve which in its general trend agrees with the experimental curve , for it is tangential to the axis of at th ' origin , and is asymptotic to the saturation line .
Unfortunately , however , for the capillary theory , if we calculate the radius at any pressure at all removed from saturation , we a value which is of molecular dimensions , and at which all ideas of surface tension become unsuitable .
Thus for a pressure of 90 per cent. of the saturation pressure the radius at ordinary temperatures comes out to be about cm .
Vapour Prcssure from Solutions.\mdash ; It is of interest to examine the pressure of the vapour of water from solutions in liquids which elves have practically no vapour pressure , to see if similar relations hold to those we have seen to hold for wetted solids .
; Regnault 's results for solutions of water in huric acid , it will be found that similar relations do approximately hold .
In Curve 4 is plotted the isothermal at C. for water and sulphuric 304 Prof Trouton and Miss Pool .
Pressure [ Nov. 30 , acid , that is to , the vapour pressure at from solutions of various strengths .
The similarity of the general trend of this curve to that obtained with flannel will be noticed .
.\mdash ; Isothermal at C. for Water in Sulphuric Acid .
In the case of the solid the curve meets the saturation line at a finite distance , whereas with the liquid the approach is asymptotic .
This variation in character is to be attributed no doubt to the constraint introduced by one substance being solid , SU that with farther accretion of water a distance is reached at which the particles of added water are outside the range of molecular attraction of the solid .
On the other hand , owing to the freedom of diffusion , this constraint is not present with liquids .
The isoneres for water vapour from solutions of sulphuric acid and watel are shown in fig. 5 as dotted lines .
The ygrometric curves which start from the same points at the temperature of C. are also shown .
These are the curves drawn with fulJ lines .
Inspection that , as in the case of flannel , the two series of curves may bc safely assumed to coincide , at least within the range of temperature which of the .\mdash ; At present there no case in which we ]lave data to enable us to draw the isoneres at relatively temperatures .
rometrics can be drawn up to the critical npcrature , beyond which point they cease to have meaning .
It is obvious that in general the isoneres must extend beyon this temperature .
That is 1905 .
] in Holding Moisture .
.\mdash ; Isomers for Water and Sulphuric Acid .
to say , a cve could always be drawn giving the pressure at tempel.atures above the critical necessary to condense or keep from evaporating a definite of water on the given material .
We have little information as to the condensation of water on surfaces at temperatures above the critical , but we kJlOW that similar condensation occurs with other substances , for instance , air and other ) ases on glass , charcoal , etc. , at ordinary temperatures .
We with probability an idea as to how the isoneres may run at temperaturcs above the critical from Henry 's law of the absorption of gases by liquids , namely , the volume absorbed is constant .
Thus with rising temperatures the pressure is proportional to the absolute temperature .
If 306 Prof. Trouton and Miss Pool .
) Pressure [ Nov. 30 , this is true the case of gases and condensing on solids , the isoneres at teml ) eratures above the cliti run in straight lines sing through the origin of co-ordinates .
The exact the isoneres take is to say without further data th are at present to hand , but we may with fair probability , as indicated ( , sketch out their general character .
This has been done in for watel on a solid .
The to } ) line is the vapour pressure up to the criticnl temperature .
curved portions of the three beneath are rometrics ( and the saturation pressure .
the coincidence of the isoneres with ] these , and in addition the extension of Henry 's law , we have the complete isoneres as shown .
FIG. for Water pour .
If we follow one of these low temperatures upwards , we see at once the enormous pressures necessary in order that a given surface may retain its layer of condensed lnaterial intact as the temperature rises .
A mate1ial such as charcuaL which condenses gases freely , ] hold a certain of gas at any temperature and its corresponding pressure as 1oy its isoneres .
If one follows an isonere from low to high enlperatures , one can tracc1 the charactel ' of the action of the charcoal in Sir James } method for obtaining ] vacua by means of condensation in charcoal .
It will be 1low seen how a weight of gas which would require an enormous tu it to condense a given surface at high temperatures , can be held by the same surface at low temperatures at an exceedingly low pressure .
1905 .
] in Equilibrium with ces Holding Moisture .
of Solid.\mdash ; When the ison eres continue beyond the critical temperature , as is the case with air on , the attraction between the particles of the solid and those of the liquid must be supposed greater than that between the ) articles of the liquid itself .
To take a simple case , first a drop of the liquid in with yapour .
The pressure cannot be lowered without evaporation taking place , but if we now suppose the part of the drop replaced by , say , flass , leaving only a thin layer of liquid of a certain thickness , the pressure can be reduced , without evaporation occurring , to a definite value depending on the thickness of the layer liquid left .
Besides doing this we may pass along the to temperatnres beyond the critical at which the material unaided by the solid could not exist at all in the liquid state .
Some substances , for example resinous solids , may ftct the other way in this respect , requiring a greater pressure in order to retain a layer under the above circumstances .
of Surfaccs Reduction in .\mdash ; The relation as found by experiment between the vapour pressure and the weight condensed on a given surface , namely , the parabolic law , us for the rate of loss of moisture from a given surface with reduction in , where is the weight held at saturation .
That is to say , the rate of loss with reduction in pressure is inversely proportional to the extent drying has proceeded .
Looked at from this point of view , experimental relation seems a suitable and likely one .
For as drying proceeds the aver distance from the solid to the particles becomes less .
The effect of ] ] attraction of the solid in retaining the layer should thus be felt in incre ratio .
If the law were known connecting the distance apart with the attraction between particles , that is to say , the Laplacian attraction , it be possible to deduce the relations of the vapour pressure and the thickness the condensed layer .
Conversely , now that we know this relation experimentally , it ests a possibility of determining the form of the function assumed by Laplace .
For example , let us make the supposition that inside the boundary of the liquid vHpour surface the total pressure is always to be the same , that is to say , the sum of the intrinsic pressure and of the vapour pressure for every thickness of water layer is to be a constant quantity .
, where , is the saturation vapour pressure , and the intrinsic due to an unlimited depth of water under the VOL. LXXVH .
308 Prof Trouton and Miss Pool .
Pressure [ Nov. 30 , , while is the vapour pressure for a thickness of water layer ( l ) , and intrinsic pressure partly due to this water layer and partly due to the solid beneath .
We the notation employed by where and are the functions proper for water-water and waterglass respectively , and " " densities\ldquo ; of water and glass .
The of the layer of water in with saturated vapour is probably in alnomt ) that of the " " molecular range so that for lowel the solid must lIave an important effect .
Parks , however , gives the thickness of water layers on lass and other substances to be from 13 to , which is greater than the commonly adopted value for the molecular it is that Quincke estimates the ntolecular to be oi about the same or cm .
The method adopted Parks makes it possil ) that some of the water supposed by to be in the watel layer was held in capillary cavities of the character eferred to earlier in this paper .
write This es us AssnIning our relation , where is the nlaxirnum thickness reached by the layer at saturation , and is a constant , we have , if we take .
is the attraction per unit mass at a point situated at distance from the plane surface of a substance of ' density\ldquo ; extending to infinity .
Thus on the supposition made the law of force between two articles , as assumed by Laplace , must be such as to produce a resultant force at distance from an infinite plane boundary of the substance , when is the molecular distance at which the force is insensible .
of Action of \mdash ; The law of the coincidence of the , ( isohygrometrics throws on the working of the various it art .
" " Capillarity ' il .
' vol. 5 , p. 517 .
1905 .
] wiih Holding Moisture .
form ) of hygroscopes , in which the alteration in of hair or employed .
Thus these substances are wetted to the same extent for the same rometric state , so that , ecting temperature effects and the length to depend simply on the wetness , we can understand how these instruments afford approximately correct indications of the hygrometlic state at all tempel.atures .
The author has utilised directly the weight of moisture condensed on flannel for the purpose of constructing a recording hygrometer .
The flannel is suspended on a recording balance .
Its alterations in weight are thus .
The amount held by the flannel alters with the temperature the moistul.e in the atmosphere to remain constant , or sllpposing the ture constant it alters with the amount of moisture , but , as we seen , both may alter without changing the weight , provided the ation is an isohygrometric .
of of Wntcr in with cretted ?
rith Constant of Wate The of apparatus shown in fig. 7 was used by us in our experiments to investigate the relation between the ture and the esponding vapour prcssure equilibrium with a fabric wetted with a constant quantity of water .
The nlaterial used was flannel , and the results obtained show that , at all temperatures within the limits of the experiments , the of the yapoul from it , when holding a constant weight of water , may be taken as being always the same fraction of the saturation pressure of water vapour .
The details of the apparatus are as lollows : is a glass bulb of about 300 capacity containing the matel'ial under amination .
is a side tube leading to ( he .
pump with which the apparatus cuated .
It }}as sealed off at the onstriction when there a vacuum .
a sidc ) to this provided with a tap , an 1 enling in a bent-over sealed capillary .
This served for the entrance of water .
10 Prof Trouton and Miss Poo ] .
[ Nov. 30 , is ( the ( in a lile ) ) which served as a lough reference } .
The ) ected to a mercnry manometer , the mercury level ) of a mirror at the } of the eservoir , fine cratched across it .
This line is made to coincide the reflection } level with the mercury level itself .
The scale is then The cnts ( ( .
8 ) wele as follows : bulb and the stem to below the mark were in a copper water ) , a rubber ( fitted into a ) the bottol n the rest of the ( of which the to top ) to project below from the .
The level always kept ) to the ; this the ) the salnc ) , so that ) lion on the mclcury ite sides of the water ; ( electric oiled ) 1905 .
] librium with Holding Moisture .
the merculy surface visil ) through these windows .
The horizontal cross wire of a telescope was focussed on the barometer tube at the level of the hair hair served as the coarse , the cathetometer as the fine adjustment .
means of the cathetonleter also this level was read on the scale the pressure auge , so that pressnles could be read by reading the position of the surfacc of mercury in the reservoir , since during an experiment the ) level of mercury was kept constant .
, to any desired temperature was cffected by lleans of stealn , which helped also to stir the water .
To aid keeping the bath at any articular t(-shaped burner placed beneath was als used .
Cooling could be cted by off hot water by a siphon and letting in cold from a tap .
The bath was kept stirred , and of the pressure taken when the telnpelature was steady at any pal'ticulal .
point .
The material used was about 10 borammes of freshly washed flannel .
This was cut into strips and inserted into the bulb .
The method of introducing the water was as follows : capillaly from tube was sealed off and the tap on .
The apparatus was then evacuated ) the pump , which was then sealed off .
After turning off the tap of tube the end of the was broken off umder the surface of some freshly boiled distilled water in beaker .
Thus the tube , as far as the tap , was filled with water , and the required amount could be into the paratus without letting any air in .
After it had evapor ted into the bulb , the side tube was taken off at the constriction near the bulb .
the apparatus was ready for experiment , but it will only serve to give one isonere or curye with the of water present .
Between each set of experiment } } it bas to be econstruc ) , freshly cuated , and a fresh amount of water let in .
The evacuation takes ) time , as the air has a tendency stick in the ] ) of the flannel .
In several experiments a rection for air to In addition , correction for capillality was required owing to the size of tubc employed .
The chief source undoubtedly was uncertainty as to temperature of the flaunel itself .
Difierent values for were got according as ture w reacbed from above or below .
The following ) the results ) sel.ved when the flannel contained nearly its maximum quantity of -ater .
In column 1 are the temperatures , and in 2 the corresponding ) pressures .
The ] ) saturated from 's tables iven in cohmtll the ratio of the pressures from tltG flannel to is iven in column 4 .
It will be seen that ratios are approximately considering the difficulty of the teml ) erattlle s for equili312 Prof Trouton and Miss Pool .
Pressnre [ Nov. 30 , brium to obtain , the ( with the suggested law of constant ratio may be looked upon as sufficiently proved .
Table I. P. These results are perhaps most conveniently examined when hibited as in .
Here the top curve is that of the saturated vapoul pressure .
second is the obtnin the ordinates of the in a constant ratio , namely , that of the of those in the last column in Table 1 .
The points lnalked along the are hose in lumJl 2 , and the whole follow the curve .
In II are iven the valions 1yhell the flannel contains a less weight of water .
Table I I. j ) P. 352 352 352 352 352 352 352 P. P. 57 .
24 .
bl 2 372 5 64 .
0.52 third CU1VC in fig. 9 the drawn the mean ratio obtained in Il , the 1 ) as ) ) eing marked it .
The observations with still less water present in flanncl .
1905 .
] in Equilibrium with Substances Holding .313 C. Mr. H. Stansfield .
Observations and [ Jan. 3 , TIT .
The lowest curve is the rometric 1 the mean ratio of III .
Tn this case the reement of the points with the curve appears remarkably , considering the experimental difficulties .
and of and By , , Pesearc Fellow of the University of JMan ( Communicated ) Arthur Schustcr , cceived January 3 , \mdash ; Read 1906 .
[ PLATES 2\mdash ; 3 .
] 1 .
olnc y , while student , in continuation inold a nlitdc measurements on a of black films .
Th of the ) ents were not for thinner ) ) lack , and I did llot n it , I it had ) not seen the also ined electrical ) sel.vationH the ) thic 1 ol ' othcl csultb of the the in subject , I wished to this ) , ined f ' il .
ans , vol. 184 , ) } .
il .
vol. )

