]\gt ; PHILOSOPHICAL TRANSACTIONS .
\mdash ; I. On the Ratio of the Specific Heats of the Paraflins , their Monohalogen J. W. CAPSTICK , M.Sc .
( Vict , Scholar and Coutts-Trotter Student of Trinity College , mbridge .
Communicated by Professor J. J. THOMSON , F.R. Received May 25 , \mdash ; Read Jnne 15 , \mdash ; Revised October 21 , 1893 .
TABLE 0F CONTENTS .
Page .
S1 .
Introduction .
THE experiments to be described in the present paper were undertaken in the hope of obtaining data which would throw light on one of the most obscure points of the kinetic theory of gases , namely , the distribution of energy in the molecule .
The properties of gases on which the kinetic theory gained its reputation were the constancy of the-product of pressure and volume , and the uniformity of the coefficient 15 .
2 .
94 NDCCCXCIV .
of expansion .
For the explanation of these in the case of the hypothetical perfect gas , no knowledge of the special constitution of the molecnle is required , but for most other properties , and especially thermal properties , the kinetic theory fails to explain the facts from want of information concerning the dynamical peculiarities of the nlolecules of different gases .
From the ratio of the two specific heats of a gas we can calculate the relative rates of increase per degree rise of temperature of the energy translation of the molecule as a whole , and the enel.gy due to the motion of the atoms relatively to the centre of gravity of the molecule .
If is the ratio of the rate of increase of the internal energy to that of the translational energy , we have the well-known equat on\mdash ; where is the ratio of the specific heats of the gas .
Thus the constant 7 has a high theoretical value as leading directly to a fundamental dynamical property of the molecule , and a knowledge of its value for a large number of gases suitably chosen would not improbably afford material on which to base a theory of the configuration and motions of the atoms in a molecule , or would at least g ) data by which to test theories based on other considerations .
Stated briefly the following is the present state of our experimental knowledge of the ratio of the specific heats .
Almost all the older work was rendered valueless by RoNTGEN'S showing Annalen , .
great an effect the size of the apparatus has on the results .
His own values for and carbonic acid are probably near the truth , but the difficulty he experienced in finding a suitable pressure gauge , and the large size of the apparatus , have caused his method to be put out of the field by KUNDT'S Dust Figure method ( POGG .
'Ann vol. 127 , p. 497 , and vol. 135 , pp. 837 and 527 ) .
The earliest experiments by this latter method are those of KUNDT and WARBURG POGG .
on Mercury Vapour , by which it was shown that is zero for the mercury molecule , and hence the molecule has no power of absorbing internal energy , thus confirming the chemical view that the molecule is monatomic .
Next we have the work Carbon Monoxide , Carbon Dioxide , Nitrous Oxide , Ethylene , and Ammonia , by WULLNER ( WIED .
'Ann vol. 4 , p. 321 ) , who , using KUNDT'S earliest single-ended form of apparatus , found that with the exception of air these gases all have ratios of the specific heats that fall considerably with rise of temperature .
Up to this time it was thought that all diatomic gases have equal to .
To test the point further STRECKH : R ( WIED .
'Ann vol. 13 , p. 20 , and vol. 17 , p. 85 ) investigated the halogens and their hydracids .
He found that hydroch]oric , hydro bromic , and hydriodic acids have the value , but that the simple halogens and OF THE PARAFFINS AND THEIR MONOHALOGEN DERIVATIVES .
iodine chloride have values near ratios of the specific heats of all seven gases were found to be unaffected by change of temperature over a wide range .
BEYME ( ' Beiblatter , ' vol. 9 , p. 503 ) made some experiments on the saturated vapours of ether , carbon bisuIphide , chloroform , benzene , and water , by a modification of KUNDT'S method , but as he made no attempt to determine the densities of the vapours , his work does little more than show that sound is conducted freely through saturated vapours .
P. A. MULLEn WIED .
investigated the ratios of the specific heats of a large number of gases by a method devised by ASSMANN .
' Ann. , ' vol. 85 , p. 1 ) .
MULLER assumes that alternate compressions and rarefactions , with a period of half-a-second in a globe holding about a litre of gas , are adiabatic .
In the light of the work of RoNTGEN and KUNDT on the effect of the slze of the apparatus , it is evident that this cannot be the case , and we might expect MULLER'S results to be too low .
In almost every case where comparison is possible his result islower than that obtained by methods recognized trustworthy .
The experiments of JAGER WIED .
were intended to test the question whether depends on the degree of saturation of the gas or not .
He concludes that for the vapours of ether , alcohol , and water the degree of saturation has no effect on , but the experiments are hardly accurate enough to be conclusive .
Other papers on single gases are those of KAYSER ( WIED .
'Ann .
, ' vol. 2 .
p. 218 ) , on Air , of MARTINI Revist .
Scient .
, on Chlorlne , and of E. and L. NATANSON , on Nitrogen Peroxide .
It appears that the gases hitherto investigated have not been chosen with a view to elucidating the constitution of the and are not suitable for this purpose .
Almost all are inorganic gases which , it is true , are easily prepared fairly pure , but are .e too irregular in their properties to lead to much of theoretical value .
Each gas has peculiarities of its own which are not shared with others , and we have nothing corresponding to the homologous series of organic chemistry .
It can hardly be doubted that the success of physico-chemical methods of late years would have been much less striking if inorganic bodies only had been available .
Amongst the carbon compounds we have many series of gases or volatile liquids proceeding by regular increments of to the molecule , the members of any one series showing such striking similarities in their properties as to point to similarity of constitution of the molecule .
We have , too , the advantage of accurately determined graphic formulae , and though we are not justified in regarding these as concrete representations of the molecule , yet the consistency with which the system ofnotat on has been applied to thousands of compourlds shows that it has its basis some physical fact , and makes it well suited to serve as the ' independent variable expressing other properties as functions of the complexity of the molecule . .
For these reasons I have chosen the parafflns and their monohalogen derlvat as being simply related to each other , easily volatile , and stable .
The method adopted for the determination of the of the spe.cific heats was KUNDT'S velocity of sound method It has the .
geto im urities ; but this is of the density of the gas to be known , and hence being very bl counterbalanced by our knowing from UNDT nves from KUNDT'S investigations all the conprobably count ditions on which accuracy depends .
Most of the gases used diverge considerably from agreement with BOYLE'S law , and have not had their vapour densities determined except by the rough methods use methods used in lae and even if they had , it would be unsafe to trust th fixing molecular formu ae ; an even ' boilin oint results , for the usual test of the purity of organic liquids , constancy il lead to erroneous conclusions , as will be seen by the work on ethyl bromi may ea described below .
To avoid error rom from this source a direct experimental determination has been made on the compounds as they were used in the velocity of sound experiment .
The formula that has been used by most investigators for calculating the ratio of the specific heats from the velocity of sound is where ratio of the specific of air , specific gravity of the gas referred to air at the same temperature and pressure , wave-length in the gas , wave-length in air .
This formula is only true for a perfect gas , for the square of the velocity of sound is only if is a constant at any one temperature .
In the present work I have used a formula obtained as follows:\mdash ; The equation , where the symbols have their usual meanings , is true for any homogeneous substance .
( RAYLEIGH'S Sound , S244 .
) From this we have But the differentiation being at constant temperature .
Hence and . . . . .
( 1 ) .
OF THE PARAFFINS AND THEIR MONOHALOGEN DERIVATIVES .
Hence , neglecting the square of , we have ( 2 ) .
Equation ( 1 ) is quite general , but the assumption that is small , limits ( 2 ) to gases .
To find the value of the last factor , I have determined imentaily the vapour densities of the gases at various pressures , and plotted a curve connecting and the sJope of which , at any point , gives the value of at that point .
It is , of course , of consequence what units are used , as the dimensions of a.re those of a number .
The formuh can be put in various forms , but that iven above seems to lead most directly to the required result .
It is not , however , easy to determine the correction very accurately , for the variation of pv is not rapid , and a small error in the density observations , or in the drawing of the curve , may make a considerable change in the slope of the curve .
For most of the gases that I have used , the correction is from one to two cent. of the whole value of 7 .
WULLNER , STRECKER , and others used the uncorrected formula , and , though they worked on gases with low boiling points , their results would be quite appreciably raised by the correction .
If there is any impurity present in the gas , will be the specific gravity of the mixture .
In the case of Marsh Gas and Ethane , this was got by calculation from the analysis of the gas , and in the case of the rest of the compounds by direct experiment .
The result is the ratio of the specific heats of the mixture , and requires a further correction if the of the impurity is not the same as that of the gas undel investigation .
A formula for effecting this may be obtained in the following way , the gas being assumed perfect as the correction is small:\mdash ; Let total kinetic energy of translation of all the molecules in unit mass the gas .
\amp ; c. , the same for each of the components .
, the pressure of the mixture .
\amp ; c. , the partial pressures of the components .
ST , the increment of ; for a rise of temperature , S. , the increment of the internal energy of the molecules ; for a rise of temperature , , the specific heat of the mixture at tant pressure .
volume .
, , Then , being the same for each component .
But for unit mass of a gas .
Therefore Therefore S2 But , since the average kinetic energy of a molecule is the same for each of the constituents , and the pressure is proportional to the number of molecules , or , / p , / P , \amp ; c. Therefore S2 ST .
Similarly S2 ST .
Therefore and since , for a single gas , the above reduces to . . . . .
( 3 ) .
This equation is equivalent to , and merely expresses the fact that the total increment of energy per degree rise of temperature is equal to the sum of the increments for each of the components .
Analyses of the marsh gas and ethane used showed that there was always a little air present .
The correction for this , calculated from ( 3 ) was only one or two parts in a thousand , which is within the errors of observation , so that for the other gases , where nothing was known as to the nature or amount of the possible impurities , no appreciable error is likely to have resulted from omitting it .
OF THE PARAFFINS AND THEIR MONOHALOGEN DERIVATIVES .
A point requiring some consideration was the question at what temperature the experiments should be made .
According to WULLNER carbonic acid , carbon monoxide , nitrous oxide , and ammonia have values of the ratio of the specific heats which change in some cases by as much as 4 per cent. between and , and if this were so in all cases , it might well be asked at what temperature the results would be comparable .
There are many gases , however , for which is constant ; oxygen and nitrogen are such , and STRECKER showed that over a long range of temperature the change , if it existed at all , was very small for the halogens and halogen acids .
MULLER , too , found no indications of in the gases investigated by him .
His assumption that the compressions and rarefactions in his apparatus are adiabatic is so improbable that we are bound to suppose his results are too low , but the method should be capable of showing relative changes .
In fact , no observer but WULLNER has ever found any appreciable change of with the temperature , and it is possible that the Tangement of apparatus at least exaggerated the change he found .
Hence , so far as previous observations go , there is a presumption in favour of the constancy of Independently of this , there is something to be said in favour of choosing some constant temperature , as the chief interest of arises from its relation to the internal energy , it seems desirable to secure that either the internal , the translational , or the total energy should be constant , and we can the translational energy constant by working at a constant temperature .
Consequently it was decided to work at the temperature of the room .
S2 .
The Kundt The apparatus used for the determination of the velocity of sound in the gases was in all essential features the same as that described by KUNDT in PoGGENDORFFS Annalen , ' vol. 135 .
The double apparatus was used , as it makes accurate temperature observations unnecessary , the tubes containing air and the gas under invest gat on lying side by side .
It also ensures the figures in air and in the other gas corresponding to exactly the same note , so that change of pitch in the vibrating tube from change of temperature or any other cause has no effect .
It will be sufficient to describe one end , as the two are almost identical in arrangement .
The vibrator , AB ( see fig. 1 ) , is a closed glass tube 150 centims. long and 35 millims. in diameter , and was chosen from a considerable number tried as givmg the best figures .
It is not desirable that it should give a very loud tone , for this scatters the dust too much , but it should speak readily , so that the intensity can be adjusted .
An important point is to choose one that gives a note as free from overtones as ' for these injure the sharpness of the figures very much , and make them Covering one quarter the length of this tube is a difficult to measure .
htl wider one , .
The joint at was first made according to 'directionS , by wrapping a strip of thin india-rubber lnany times round , and wlrlng it , but this proved very unreliable .
It requires a great deal of care to such ajoint even approximately etting leaky and requiring to be patched up with , and it is contlnua finally , I had made a wide tube with thick walls of the best soft ubber solution , so , on slipping this on and wiring it down a perfectly tight joint was made rubber , an The sam6 rubber tube was use was for connecting CE with the semi-circular copper , thus making a flexible joint and the conduction of the sound through the walls of the tube The tube , in which the dust figures are made , is 125 centims. long , and 26 millims. in internal diameter .
At the end , , a brass union is fixed on withsealing-wax , and by screwing up tightly the two brass faces with a lead washer between , a joint is made that is air-tight , but can easily be taken apart to measure the figures and put in fresh dust .
At the other end , , is a similarly detachable cap bearing a stufling box .
Through this passes a narrow brass tube with a disc on the end , , by means of which the vibrating length of the column of air can be varied , so as to give the best figures .
For apparatus such as this , lead glass seems to be much better than soft German .
It impossible to put on the caps without some strain , and with the German glass much trouble was caused by the tubes breaking at awkward times and wasting precious gas .
Since lead glass has been put in , there have been no breakagea Through the tube connection can be made with either a Topler mercury pump OF THE PARAFFINS AND THEIR MONOHALOGEN DERIVATIVES .
9 for exhausting the apparatus , .
Sprengel for extracting a sample of the gas for analysis , or a water pump for regulating the pressure of the contents .
At is a side tube connected through a pressure gauge with the gas holder and drying apparatus .
With the connections made as thus described , there was found to be very little leakage in the apparatus .
When it was exhausted as completely as possible the rise of pressure was only a very small fraction of a limetre per hour .
Different methods of filling were adopted according to the ma , terial used .
the earlier experiments a water pump only was used .
By means of this , the apparatus was exhausted to from 15 to 20 urillims .
, and the gas admitted slowly through the purifying train , the process being repeated several times ; but the method was too extravagant for the more costly materials , and took too much , so , the later experiments , a ToPLEB pump was used .
This has a large and with an hour 's pumping the pressure was reduced so low that it was often difficult to see whether the gauge or the barometer by the side of it stood the higher .
When the gas was admitted through purifying and drying apparatus , this ratus was usually exhausted with the rest , but as in most cases GEISSLER bulbs were used , the vacuum got gradually worse in the successive bulbs from the pressure of the contained liquid , so they were first filled with the gas by exhausting them two or three times and allowing it to stream in .
When a volatile liquid was used , it was contained in a small bottle with a tight cork , through which passed a glass tube , and wired on the end of this was a piece of thick-walled india-rubber tube .
Before attaching it to the Kundt apparatus , the liquid was made to volatilize freely by warming it , or by connecting the bottle to a water pump , and when the was driven out the india-rubber tube was closed with a clamp , and joined to the dust tube , so that when the exhaustion had been completed , by opening the damp , the vapour could be admitted free from air .
For the hydrocarbons and methyl and ethyl chlorides , lycopodium powder was used in the dust tube , as it gave decidedly better figures than silica , but with the dense gases it was found to become sticky , so that it could not be made.to moye , and for these silica was used .
A preliminary experiment was always needed to fix the position of the piston , for though with the denser gases figures of some sort could be got with it in any position , yet they were generally unsymmetrical unless it was carefully placed .
The position which gave the strongest agitation gave the most symmetrical figures , but they were seldom perfect in .
respect , and for reason the proper distribution of the dust in the tube is a matter of importance .
There is usually a tendency for the powder to encroach on the node from one side more than from the other , and this is more marked the greater the quantity of dust used , so that , if the quantity per centimetre varies from end to end of the tube , the result is an apparent shifting of the nodes to an extent that is not the same at rent po nts .
If the dust is uniformly spread , this want of symmetry has no influence on the measurements .
for utting in the dust was to draw through the tube a The method I employ or dry doth which.deaned out what was left from the erlment and slightly electrified the glass .
The tube was then pace a gp 1 in a in osition , and powder in at the end through a small and aowe to run th in the llowed to run through gently in a narrow stream .
It was then turned end for end , and the dust pour roug direction , the result of which was that the small fication cause as muchopp to adhere to the glass in a narrow uniform strip as served for the purpose of .
the experiment .
When it was placed in position and the were to be made the tube was turned round , so that the dust did not lie along the bottom but was a little way up the side ; then , on rubbing the vibrator with a piece of wet flannel , the powder ran to the lowest point everywhere but at the nodes , which were left as clear spaces , narrow and sharply bounded , separating rectangular patches of dust of great ularity .
Fig. shows part of a set of figures obtained with isopropyl bromide .
Fig. 2 .
For the measurement of the figures two pal.allel platinum wires were carried on a framework sliding .
a steel scale divided to millimetres ( fig. 2 ) .
These wires were placed so that the tube lay between them , and their plane passed through the centre of the clear space at the node , and the position of the framework was read on the scale , tent of a millimetre being estimated with the help of a lens .
When the figures were of average quality the setting could be repeated so that the positions did not vary by more than two or three tenths of a millimetre .
Table I. gives a typical set of measurements .
They were made in one of the methyl bromide experiments , and are chosen as being neither the best nor the worst of the sets , but a fair average .
The first column gives the scale-readings , and the second column the half-wave lengths got by subtracting the consecutive readings from each other .
The first half- OF THE PARAFFINS AND THEIR MONOHALOGEN RlVATIVES .
dozen figures next the vibrating tube are omitted .
This was generally done , as they were almost always found to be irregular and less distinct than the rest .
TABLE I. The mean value for the half wave-length is , and it will be seen that no single measurement differs from this by more than six-tenths of a millimetre .
In two or three sets where the figures were poor divergence from the mean reached as much as a -millimetre , but was never greater .
In some of the propyl chloride experiments it was not more than a quarter of a millimetre .
The method of calculation of the mean was to divide the , of the nodes into two equal sections , subtract each reading in the first section from the corresponding one in the second , take the mean of these differences and divide by the number of half wave-lengths between the first readings of the two S3 .
The Vapour Density Apparatus .
The ordinary methods for determining vapour densities are not very for an investigation such as this .
HOFMANN'S , VICTOR MEYER'S and DUMAS ' are scarcely accurate enough , and the two latter are not applicable without modification to pressures other than that of the atmosphere .
REGNAULT'S , though very accurate , would take too much time when so many determinations have to be made .
As the vapour densities are only required at the of the room , he conditions are much simplified , and I have devised a form of apparatus , using HOFMANN'S principle , which gives results concordant to per cent. without an .
great expenditure of time .
Doubtless with greater precautions for securing uniform tie of temperature higher accuracy might be obtained , but ) error of one part a thousand is well within the experimental errors of the of the work .
AB is a glass tube centimS .
long and millims. in diameter ( see fig. 3 ) , at A and sealed at to the curved tube , the middle of is straight and horizontal .
Before sealing to the tube , the latter is calibrated , and the volume determined between the end and a file mark , near the upper end .
CD then attached to , and the of the etermlned by filling with water and weighing .
from this the volume of DK we get the of , and as tube CD has been already calibrated the volume to any uher point is known if required Fig. 3 .
Next the side tube , , of the same bore as , is sealed on , and the three-way tap , , making connection with an air pump or with the mercury reservoir , H. is a thermometer graduated to fifths of a degree .
A small quantity of the liquid whose vapour density is required is sealed up in a small tube with capillary ends and weighed .
This is introduced at , and made to rest at , by inverting the apparatus for a moment .
Next , is dosed with an bber stopper , being also closed , whilst the apparatus is exhausted through the three-way tap , , after which operation is turned so as to allow the mercury to flow in from the reservoir , and is opened .
The difference of the levels of the mercury in the two tubes is read by means of a cathetometer , and this difference subtracted from the height of the barometer gives the pressure of any air left iu the apparatus .
The small tube is then broken , by tilting the apparatus a little and allowing it to slide over into AB , where the capillary end breaks off and aJlows the liquid to evaporate .
By reading the levels a second time we get the pressure of the vapour , and OF THE PARAFFINS AND THEIR MONOHALOGEN DERIVATIVES .
knowing its weight , volume , and temperature , we have all the terials required for caloulating its specific gravity .
It is not advisable to exhaust the apparatus very in the first instance , for then the evaporation of the liquid is so violent that fragments of glass , and sometimes even the whole tube , get blown over the bend at the top on to the surface of the mercury , making it difficult to read the position of the surface .
If 15 or 20 millims. of air is left in seldom happens , but time must of course be given to -allow the gases to diffuse into each other .
To avoid draughts and inequalities of temperature , the whole apparatus , with the exception of the reservoir , , is enclosed in a box , with vertical openings at the front and back , through which the levels can be read with the cathetometer .
The calculations are simplified if the reservoir is always adjusted so that the level of the mercury in CD stands at the same point .
This makes the pressure of the residual air the same in the two measurements , proyided the temperature is constant .
In my experiments I always brought the level to the file mark , K. To empty the apparatus the reservoir is lowered till the mercury in CD sinks to the level of G\mdash ; E bein of course closed\mdash ; and air allowed to enter through the three-way tap .
The vapour densities were generally required in the neighbourhood of certain determinate pressures .
To secure this the liquid was always sealed in a tube of the same diameter , and a preliminary filhng and weighing being made with water , simple calculation gave the length required in any case .
It was then easy to draw off a piece of a length as would hold within 5 or 10 per cent. of the requ re red amount .
When the liquid a very low boiling point special arrangements are needed for filling and sealin the tubes .
The following method has been found to be qu te satisfactory , but requires careful manipulation to avoid breaking the capillaries .
A piece of glass tube is drawn out into the form ABCDE ( fig. 4 ) , with a capillary part at , and a end , DE , and weighed .
The end , , is then connected with a water pump , by means of a pece la india- rubber tube closed with a spring clamp .
The part , which is that which is to be filled , rests in a lead tray , slightly inclined , and filled with a suitable freezing illary end , DE , dips the level of the liquefed gas , which mixture , and t , he cap is contained in a tube surrounde a freezmg mixture .
By openmg the clamp for a moment the liquid drawn into and the capillaries sealed off by a , at , On weighing with par he arts drawn off we get the amount of ] iquid eosed .
Methyl chloride requires a temperature below to liquefy it , and , for this , ether and carbonic acid is most .
convenlent , but has the disadvantage of glVlng off an inflammable vapour , take fire from the flame Hence the tube was packed round with solid carbonic acid , moistened with chloroform , forms quite as effective a freezing xture and does nob readily take fire The condenser , , was closed with a stopper , through which passed a tube to carry the ether vapour beyond the reach of danger S4 .
Marsh Gas .
As was to be expected this gas gave much more trouble than any of the others .
It cannot be freed from air by liquefaction , as was done with most of the others , and as the density of the gas is a factor in calculating the ratio of the specific heats , was necessary to make a set of analy@es after each experiment to determine the percentage of air .
The correction for this is by no means inappreciable , on account of the low density of methane ; roughly speaking , one per cent. of .
air makes an altelation of one per cent. in the result .
In the case of such gases as t , he paraffins , the quantity which can be taken for analysis is so small that any error in its measurement from want of accuracy in the ation of the measuring tube , or other causes , has a large effect on the calculation of the percentage of air .
Adding to this all the } , sources of error incidental to gas analysis , such as incomplete combustion , oxidation of the nitrogen present , errors , \amp ; c. , the result is that the accuracy is less than that attainabl the of sound determination .
It is unfortunate too that this .
additional source of error should enter most prominently in the case of methane , which is notably a most difficult gas to prepare pure .
The consequence is that the range of values found for the ratio of the speciiic heats is greater than for any other gas , and the most that can be said is that the mean is probably within one or two per cent. of the truth .
For the preparation of marsh gas two methods were used , FRANKLAND'S method by the action of zinc methyl on water , and GLADSTONH : and TRIBE'S by the action of the copper-zinc couple on methyl iodide and alcohol .
The latter method appears simple when the original memoir describing it is read , but in practice it requires considerable care .
It would be tedious to recount the discouraging series of failures before gas was obtained sufficiently pure for the so the conditions on which success was found to depend will be stated simply .
OF THE PARAFFINS AND THEIR MONOHALOGEN DERlVATIyES .
There must be no water left in the apparatus , or in spite of chemical equations some free hydrogen will be given off .
The couple was several times washed with alcohol , which had been scrupulously dried with lime and anhydrous copper su.lphate .
The copper-zinc couple itself seemed to be the best drying agent for removmg the last traces of water , for the apparatus gave purer metbane the second or third time of using than it did ' first time ; hence , after setting it up it is advisable to put in a little methyl iodide , and allow it to stand for a day or two with a Bunsen valve or some such arrangement attached .
The gas that comes over first is purest , so that no attempt should be made to secure a theoretical yield .
A considerable quantity of methyl iodide escapes the scrubber , .
must be removed in some way .
A set of Geissler bulbs filled with fuming sulphurlc acid was used in this and similar cases and proved quite effective .
The first bulb blackened and deposited iodine long the second was coloured , and ma , ny litres of gas could be passed through before the colour reached the third bulb .
This intro ntroduced sulphur dioxide into the gas , to remove wlJich it was collected in a gas-holder over soda solution and shaken with it .
It was admitted into the Kundt apparatus through three U-tubes , the first containing solid potash , to remove any sulphur dioxide still remaining , the second containing nine grams of palladium black as a precaution to retain any ee hydrogen , and the third containing pumice soaked in sulphuric acid to dry the gas .
Palladium is not altogether satisfactory for the of hydrogen ; it is very fickle in its action , sometimes for no obvious reason refusing to absorb it .
In the preliminary experiments and in the preparation of propane , to be described later , 30 grams of thirl foil , superficially oxidized by ignition in air , was used , but this , though quite of ffective in removing the greater part of the hydrogen , which was all that vanted in the case of propane , failed to take out the last traces ; so 9 of the foil was con verted into " " black ignited in air , and placed in a -ttlbe in boiling water , according to HEMPEL'S directions in the methane experiments .
To remove the air from the Gladstone and Tribe apparatus , a little dry adcohol was put in , and it was then connected with a water pump and warmed till nearly all the alcohol had boiled away , but the large volume of the apparatus , the great absorbing power of alcohol for air and other gases , and the long train of apparatus required , must be taken as the excuse for the large , of resen Two analyses and the calculation of the result are given in full for the first experl meant .
The first gives as half the contraction , and the formed , their ratio being .
The second gives for the half-contraction , and for the , their ratio being These ratios should be unity for pure methane .
The difference between half the contraction and the volume ofgas taken , and between the formed the original gas , gives two estimates of the air from each analysis .
These are and from the first , and 2'05 and from the second .
The discordance of these is wider than was usually obtained .
An error of millims. in reading the level of the mercury when measuring the volume of the gas taken would account for the difference .
The measuring tube of the Dittmar gas analysis apparatus was rather too narrow , as the shape of the meniscus varied with the state of the surface of the mercury .
The mean of the four gives cent. for the air .
The S.G. of the gas is got from the equation which gives Two sets of measurements of the methane figures gave as the half wave-lengths millims. and millims. , and the length of the air figures was millims. , the temperature being Hence , the ratio of the specific heats of the mixed gases is 1 .
Finally , from the equation , we have , which gives OF THE PARAFFINS AND THEIR MONOHALOGEN DERIVATIVES .
The results of the remaining two experiments made on methane prepared by GLADSTONE and TRIBE'S method are given in the second and third line of Table II .
, with the temperature , the ratio of semi-contraction to , and the percentage of air .
The next three experiments were made on marsh gas got by the action of zin methyl on water .
This method is not attractive from the offensive nature of the zinc compound and the olence of its reaction with water , but it gives a pure product .
The zinc-methyl was made 4 digesting methyl iodide with a couple on the water bath , and distilling off the product on an oil bath .
The reaction was very complete , the contents of the flask after the first operation being quite dry on cooling , but to ensure the removal of any unaltered methyl iodide a stream of carbonic acid was passed through the flask for some time whilst it was kept at In the final distillation the end of the condenser dipped below the surface of dry ether cooled in ice , by means of which loss was prevented and an almost theoretical yield obtained .
The principal reason for mixing the zinc-methyl with ether will be detailed at length in the description of the preparation of ethane , which was the first gas investigated .
What is said there will in all probability apply with even greater force here , where the compound is more easily dissociated , and the reaction more violent .
The addition of ether adds very much to the comfort of the experiment , for the mixture can be poured from one vessel to another without any greater inconvenience than strong fuming .
To prepare the methane the mixture of ether and meth was dropped slowly into a flask containing distilled water and the gas evolved collected without purification over boiled water , with which it was shaken to remove as much ether as possible .
It was passed into the Kundt apparatus through two sets of Geissler bulbs of strong sulphuric acid to remove the ether and traces of methyl iodide , one of potash to absorb any sulphur dioxide formed in the first two , and another of sulphuric acid to dry it .
After this treatment it issued without In experiment IV .
the potash was by mistake omitted , and the consequence was that the gas was found to contain.5 per cent. of sulphur dioxide .
The results of the three experiments are shown in lines III .
, , and V. , of Table II .
TABLE II .
I have no data from which to calcuJate the correction factor From the fact that the gas is at ordinary temperatures far above its boiling point , is probably ] and has been neglected .
S5 .
Methyl Chloride .
The material was made by passing a stream of hydrochloric acid gas into a boiling solution of zinc chloride in methyl alcohol , contained in a flask with a reversed condenser .
In the first two expeliments recorded below the issuing gas was passed through potash solution , and collected in a gas-holder over strong brine , as it is too soluble in water .
It was admitted into the Kundt apparatus through one set of Geissler bulbs conpotash , and one containing sulphuric acid .
As the gas liquefies it seemed desirable in order to have greater certainty OF THE PARAFFINS AND THEIR MONOHALOGEN DERIVATIVES .
of the absence of air to use the liquefied gas .
This was done in the remaining experiments .
The gas on issuing from the apparatus in which it was prepared , was passed through potash solution and sulphuric acid , a- was then condensed in a glass tube standing in a freezing mixture of ether and solid carbonic acid .
Part was then redistilled into the apparatus shown in fig. 5 , which was also used in a similar manner in the propane experiments .
A is the tube in which the methyl chloride was collected on its evolution from the apparatus in which it was prepared .
The condenser X consisted of two beakers , one inside the other , with a large boiling tube suspended in the inner one by the wooden cover .
This tube contained the freezing mixture , and in it was the small test-tube , closed air-tight by a per through which passed two glass tubes , one of them reaching to the bottom .
being closed and and open , the liquid in A evaporated off quite slowly in consequence of the cooling produced by this operation , and passing through which contained soda-lime , and , which contained sulphuric acid , was condensed in , anything remaining uncondensed paesing into the air at E. When ) tube was almost full was closed , and the tube of methyl chloride taken out of the freezing mixture , which caused it to evaporate and drive out the air from above it .
When this evaporation had gone on for a short time , was closed and opened , admitting the vapour into the Kundt apparatus .
Four determinations of the vapour density gave the following results , the pressure and temperature being recorded in each case .
TABLE II1 .
In the experiments the pressures were read to 05 millim. In this and all the following tables I have given them to the nearest millimetre .
The numbers in the third column are the specific gravities of the gas referred to air at the same temperature and pressure .
These values are plotted in fig. 6 , and from the curve the values of the density are taken for the pressures at which the velocity of sound experiments were made .
To find the correction factor , the following method was adopted:\mdash ; Taking the reciprocals of the densities given in Table , we get values of in arbi20 , J. .
CAPSTICK ON THE RATIO OF THE flEATS .
these the pressures , the corresponding volumes are trary units .
es These are given in Table IV .
, and in fi 7 the plotted on a curve , taking as The of this curve to the horizontal axis at ordinate ) as any point gives the of ab that pomt .
To get To et the correspondin to the pressures used in the of soun experiments , it is sufficient to take an approximate value of and by the pressure .
The volumes so obtained are given in the second column of Table V. , and , iding the rate of change of at these points by the pressures , we get the numbers shown in the third column .
Fig. 6 .
Fig. 7 .
TABLE IV .
TABLE V. We have then , h'nally , the following table for the ratio of the specific heats , where\mdash ; OF THE AND THEIR MONOHALOGEN DERIVATIVES .
pressure of the gas in the Kundt tube .
temperature of the gas in the Kundt tube .
half wave-length in methyl chloride .
half wave-length in air . .
of the methyl chloride at the pressure given in the first column .
TABLE VI .
In the last experiment there was per cent. of air in the gas ; the result is corrected for this .
S6 .
Methyl Bromide .
The material was obtained from KAHLBAUM , and was dried with calcium chloride and redistilled .
The results of the vapour density determinations are shown in Table VII .
and fig. 8 .
TABLE VII .
Fig. 8 .
The densities used in the calculation of the ratio of the specific heats are taken from this curve .
The correction factors are determined in exactly the same way as was lained under methyl oride .
It is needless to give the intennediate tables and curves in every case , so , for the remaining gases , I shall content myself with giving the experlmental data from which the correction was calculated , and its value .
The following table gives the final results for methyl bromide : TABLE VIII .
S7 .
Methyl Iodide .
The material was purchased from KAHLBAUM , and was dried with calcium chloride and fractionated .
Table .
and fig. 9 show the results of three vapour density determinations .
Table X. gives the final results .
the columns having the same meanings , and being obtained in the same way as before .
TABLE IX .
OF THE PARAFFINS AND THEIR MONOHALOGEN DERIVATIYES .
TABLE X. S8 .
The first attempts to prepare ethane were by the electrolysis of a saturated solution of potassium acetate .
Since these were made , Dr. T. S. MURRAY has published an account*of an extensive investigation of the method , so that it is needless to give any detailed account of my failure .
The sample first analyzed was made by using a strong current for a short time , and proved to be almost pure ethane , but as the apparatus soon began to get hot , a much smaller current was used when prepar ng a large quantity for a velocity of sound determination , and the result was that the gas was not good enough for use .
This agrees with MuRRAY'S conclusion that high current density and low temperature are necessary , and shows that the method is not suitable for making a large supply , as a strong current and low temperature are not easily secured together .
This method having failed , it was decided to use the reaction between zinc-ethyl and water .
The zinc-ethyl was prepared in the same way as the zinc-methyl previously described .
For the preparation of ethane , the zinc-ethyl was mixed with twice its weight of * ' Chem. Soc. Journ January , 1892 .
dry ether and dropped into distilled water .
The ga8 came off without undue violence , and the deposit of zinc oxide left in the flask was pure white .
The gas was collected over a large quantity ater , anfi t .
it was shaken with it to remove as much of the ether as possible , and in the rst two experlments passed slowly into the Kundt apparatus through two sets of Geissler bulbs sulphuric acid .
On opening the apparatus the ethane was.found to be without smell , but to ensure the removal of ethyl iodide which , from its high denslty , would have a very effect , in the other experiments , the gas was through one set of bulbs of Nordhausen acid , two of potash , and two of strong sulphuric acid .
As it seemed undesirable , however , to introduce ether vapour , an attempt was made to prepare the ethane by droppin the zinc-ethyl itself on ice without diluting with ether .
The result showed that the ether was necessary , for after repeated attempts , the residue left in the flask instead of being white , was always dark grey , and effervesced slightly with acid , showing the presence of metallic zinc .
Moreover , analysis showed that there were heavy hydrocarbons present , for 100 volumes of the gas gave , on explosion with oxygen , volumes of carbon dioxide .
After passing the ethane slowly through Nordhausen acid , volumes gave volumes of , so that the impurities are mainly unsaturated hydrocarbons , but probably there is some butane present .
The cause of the impurity of the gas appears to be the violence of the reaction .
The zinc-ethyl never got clear.of the dropping tube , but was immediately acted on by the water vapour , and formed a great spongy clot round the end .
This absorbed more zinc-ethyl , which was decomposed in its pores , and so the temperature rapidly rose high enough to bring about dissociation .
It is known that at a moderately high temperature zinc-methyl decomposes into zinc and hydrocarbons , and probably a similar thing happened here .
Table .
shows the results of the experiments , taking for the specific gravity of the gas the theoretical density , .
In all the experiments the pressure was that of the air .
TABLE XI .
OF THE PARAFFINS AND THEIR NONOHALOGEN DERIVATIVES .
From the approximate equality of the densities of ethane and air the presence of 1 per cent. of the latter makes a change of less than per cent. in the value of , so that even if there were as much as 5 per cent. of air in experlments III .
and IV .
, the result would be hardly affected .
From want of experimental data I have omitted the factor , but , as in the case of methane , ib probably does not differ much from unity .
S9 .
Ethyl Chloride .
The material was prepared by hydrochloric acid into a boiling solution of zinc chloride in ethyl alcohol , the resulting gas being passed through water and sulphuric acid , and condensed in a freezing mixture .
It was then redistilled , shaken with lime to remove hydrochloric acid which was still present , allowed to stand two days over calcium chloride , and again distilled through a tube of lime .
With the vapour density apparatus that I was using at the time , tubes containing euough ethyl chloride to give pressures above 450 millims. would not go round the bend at the top .
Six determinations were made at pressures ranging from 100 millims. to 453 millims. , and the rest of the curve got from the relative densities as given by a direct observation of the values of the product for a constant mass of the gas in the usual way .
A glass tube one metre long , closed at one end and graduated approxin ) ately in cubic centims. , was carefully calibrated .
After being filled with a little ethyl chloride was allowed to bubble up into it , and it was then connected at the lower end by an indla-rubber tube with another .
glass tub of the same and ope to the air .
By altering the position of this second tube the pressure on the gas could be varied , and readings taken by means of a cathetometer of a series of pairs of corresponding values of and As a test of the accuracy of the calibration and the various temperature corrections , several preliminary experiments were made on air , which made the product appear to increase at low pressure by as much as 1 per cent. After re-calibrating the tube and hunting in every direction for the cause of this , it was at last found to be due to an error in the scale of the barometer that was being used .
Making a correction for this , the of air came out quite constant , and hence it was conduded that the calibration of the tube and the temperature corrections were right .
The following table gives the details of the experiment made on ethyl chloride .
The first three columns give the corresponding values of pressure , volume , and temperature , and the last gives the quantity , which is inversely proportional to the specific gravity , referred to a perfect gas at the same temperature and pressure .
TABLE XII .
These values of are plotted on the curve in fig. 11 .
Fig. 11 .
Next six determinations of the , of the vapour were made with apparatus OF THE PARAFFINS AND THEIR DERIVATIVES .
previously described .
These had a range of pressures from 179 millims. to 453 millims. , and the results are given in the first three columns of Table XIII .
To extend the values to the higher pressures the following method was adopted .
From the curve in fig. 11 the values of were taken for the pressures at which the absolute determinations were made .
These are given in the fourth column of the table , and should give a constant if multiplied , by the numbers in the thira ' column .
The fifth column gives these products .
The extreme variation is one part in a thousand , and the numbers do not increase progressively in either direction , which shows the consistency of the two series , and is a test of the degree of accuracy reached in the vapour density determinations .
TABLE XIII .
other experiments were made at pressures below millims. , results about 2 per cent. beIow these , but the dust figures obtained were poor , and the density and correction factor had to be by extrapo lation .
As the curve is getting a little irregular , so as to make the correction term uncertam , even at the lowest pressures actually observed , it unsafe to go beyond the limits of direct experiment , hence they have been S10 .
Ethyl Bromide .
In the case of this compound the experiments were conducted in a slightly different way from those ah.eady described .
A vapour density determination was made in the usual way with a particular sample of the liquid , and when the pressure had been measured , the Kundt apparatus was filled to tffi same pressure with the vapour of an exactly similar specimen .
Thus each line in the table below gives the result of a pair of parallel experiments , the vapour density determinations not being comparable with each other , as they were made on samples of liquid which had received different treatment .
Ethyl bromide seems to be more subject to impurity than any of the other substances .
The first same used was given to me by a friend , but its vapour density was so abnormally low that I discarded it without attempt at purification , and procured a supply from KAHLBAUM .
This had a fairly steady boiling point , almost all comin over between and , and was used after a simple fractionation in the first experiment .
The vapour density determination gave a result3 per cent. below the theoretical value .
As the boiling .
point was constant and at the right temperature , this raised a suspicion that there was some impurity present which had nearly the same boiling point as ethyl bromide , but a lower vapour density .
If , as is not unlikely , the substanee had been prepared from potassium bromide , alcohol , and sulphuric acid , the impurity might be ether , so the remainder was shaken with strong sulphuric acid and redistilled , which raised the density by nearly 1 per cent. This shows the msufficiency of the boiling point as a test of the purity of a liquid , and the value of a vapour density determination as a confirmatory test .
The second and third experiments were made on the liquid after this treatment , and the fourth , after a repeat , ition of the process .
The densities obtained in this way are insufficient to give the correction factor so a Law experiment made in addition , with the following results , all at a temperature of :\mdash ; OF THE PARAFFINS AND THEIR MONOHALOGEN DERIVATIVES .
From these the correction factors were determined , the final results being shown in Table XVI .
TABLE XVI .
It is assumed , in making these calculations , that small changes in the state of purity of the liquid do not appreciably alter the relative vapour densities , one curve is used to give the correction term for all .
S11 .
Propame .
This gas was prepared by SCHORLEUNER'S method of reducing isopropyl iodide with zinc and hydrochloric aci\amp ; The presence of free hydrogen in the gas is of no consequence , as i removed by the liquefaction .
The isopropyl iodide was ade from glycerine , phosphorus , iodine , and water , us ng the proportions given by BEILSTEIN .
It boiled very constantly at 89 , showing that no considerable quantity of allyl iodide was present\mdash ; any small quantity of this , however , would not interfere , for the propylene produced by its reduction would be removed by Nordhausen acid .
To prepare propane from the isopropyl iodide , .
it was placed in a flask with granulated zinc and dilute hydrochloric acid .
The issuing gas was first washed with water , and then passed through fuming sulphuric acid to remove the isoprop.yl iodide .
Next it was passed through potash solution , and over 30 grams of palladium foil , to remove the hydrogen .
This is , of course , not essential , but prevents waste of propane in the liquefaction .
Finally it was collected in a gas-holder , over caustic soda solution , to remove any dioxide still present .
The liquefaction of the gas was carried out in the apparatus described under methyl chloride and by the same method , the soda-lime tube being omitted .
In consequence of an accident only a single determination of the absolute density of the gas was made , according to which the specific gravity lS at and under a pressure of 260 millims. To extend the result to the higher pressures a determination of the relative densities was made in the same way as for ethyl chloride .
Table XVII .
and fig. show the results : TABLE XVII .
Table XVII .
gives the results for the ratio of the ecific heats .
Th the specific gravity are taken from the curve above , an ordinate 2 sPecific eats .
values of correspond to a specific gravity 608 being taken to OF THE PARAFFINS AND THEIR MONOHALOGEN DERIVATIyES .
S12 .
Normal Propyl The material was obtained from KAHLBAUM , and was dried and re-distilled .
Almost all came over between and , the residue being rejected .
The results of the experiments are shown in the tables and curve below .
TABLE XIX .
Fig. 13 .
S13 .
Isopropyl Chloride .
The material used was obtained from KAHLBAUM .
On fractionating nothing came over below , and almost all before the temperature reached 36 .
The small residue was neglected .
Table XXI .
and fig. 14 give the results of the vapoul density experiments , and Table XXIL gives the final values for the ratio of the specific heats .
TABLE XXI .
TABLE XXII .
S14 .
Isopropyl Bromide .
The material was obtained from KAHLBAUM , and boiled ver constantly at 60 .
\amp ; the maximum vapour pressure at the atmospheric temperature is low , only a OF THE PARAFFINS AND THEIR MONOHALOGEN small range of pressures was available .
Hence all the experiments were made at 90 millims. , and instead of plotting a curve from the vapour density pressures near determinations , the mean of the three was taken and used in the calculation of The values are shown in Table XXIII .
XXIII .
These will not give the correction factor , hence a determinatio1l of the relative densities was made with the cwing result:\mdash ; TABLE XXIV .
The curve plotted from these gave 017 for ] , and using this value , Table .
gives the values obtained for TABLE S l5 .
Discussion of the Results .
Gathering the results together , we have the Wing table : \mdash ; TABLE XXVI .
It will be seen on referring back to the separate ] methyl and ethyl chlorides and a few others of the gases that the values of are slightly higher at the higher pressures .
This circumstance suggests a doubt as to the lawfulness of taking the mean , for if the change were at all considerable , the right thing to do would be to extend the range of the observations till a constant value was reached , and use this value for comparison .
The change observed can hardly be said to be beyond the range of experimental error in any case , and is perhaps only accidental .
The only experiments I am acquainted with that have been made to test the question whether varies with the pressure or not are those of JAGER ( ' WIED vol. 36 , p. 165 ) , who concluded that it does not .
His results for ether vapour show close concordance at saturation and half saturation , but the discordance of the results for alcohol and water lessens the value of those for ether .
The specific heat at constant pressure includes the change of potential energy due to separation of the molecules , and hence will probably not be quite independent of the pressure , if the gas does not obey BoYLE'S La if the change in is due only to this , it is not likely to be great .
The question has arisen quite incidentally in my work , for I was not looking for any such effect , and did not plan the experiments so as to make it perceptible .
1 have in no case used a very .
range of pressures , and have always avoided going near saturation , where the effect might be expected to be most noticeable .
The point is one that ought to be settled .
Meanwhile , the obvious law to which *See Professer lRALD , ' Roy .
Soc. Proc vol. 42 , OF THE PARAFFINS AND THEIR MONOHALOGEN DERIVATIVES .
the mean results conform , affords some justification for regarding the value of as approximately independent of the pressure .
It is plain that the gases fall into four groups , the members of any one group having within the limits of experimental error the same ratio of the specific heats .
These groups are 1 .
Methane .
2 .
Methyl chloride , bromide , and iodide .
3 .
Ethane and its derivatives .
4 .
Propane and its derivatives .
So that with the single exception of methane , compounds with similar graphic formulae have the same 7 .
Methane was almost the last gas that I ated , and it was in consequence of its appearing to away from the law , that exceptional trouble was taken to secure that it should be pure .
All who have worked with this gas know how difficult it is to prepare it free from hydrogen , and presence of hydrogen would raise the value of ; but the precaution taken of passing the gas over palladium , the concordance of the results for methane prepared by the two different methods , and the evidence of the analysis show that there could not be anything approaching enough hydrogen present to account for the difference .
Nor can we account for the difference by supposing the results for the three methyl compounds to be too low , for , apart from the fact that there are three of them , and that eir values for agree fairly well with each other , the most likely error in their case is that due to the presence of air and moisture , which would make the results too high .
Hence we must conclude that methane has not the same as its three substitution products .
It is strange that it should break through a law that appears to hold for all the other gases , but the circumstance is not without parallel .
MENSCHUTKIN'S etherification values for the fatty acids , for instance , show a similar feature , as do PERKIN'S molecular rotation constants , and the viscosity coefficients of the same series of acids .
In each of these cases a law is found to hold for all the members of the series except the first one or two .
It appears , then , that as a law to which marsh gas is an exception , one hydrogen atom of a paraffin can be replaced by a halogen atom , without affecting the of the gas , and consequently without altering the internal energy of the molecule .
This result is similar to that which STRECXER obtained for the hydracids of the halogens , for he showed that hydrochloric , hydrobromic , and hydriodic acids have all approximately the same as hydrogen .
It should however be noticed that he found the introduction of a second halogen atom caused a large fall in , the elementary gases , chlorine , bromine , and iodine , having ratios nearly equal , but much lower than those of the acids , from which fact we may anticipate a similar feature in the case of the paraffins .
36 MR. J. W. CAPSTICK ON THE ] HEATS OF THE PARAFFINS .
bstitution roducts that have more than one I am present working on the su halogen atom in the molecule , and mtend also to etermlne hither other chemically similar atoms , such as oxygen and sulphur , or car on an silicon can be interchanged thought altering the value of .
Until these experiments are finished , it be premature to enter into a discussion of the theoretical bearmg of the results .

