salts , that their specific heats may be calculated in close approximation with those found experimentally on the assumption that the constituents , anhydrous salt and water considered as ice , have the same specific heats in them as in the free state .
By the same method , ALLUARD* ( in 1859 ) determined the specific heat of napthalene .
SCHAFARIEK , lastly , has executed by the method of mixtures a series of experiments on the determination of the specific heats of vanadic , molybdic , and arsenious acids .
Quite recently ( 1863 ) , PAPE : has published investigations on the specific heat of anhydrous and hydrated sulphates .
He worked by the method of mixture , which he modified in the case of salts rich in water , by placing them in turpentine , and observing the increase of temperature produced in the salt and in the liquid by immersing heated copper .
As a more general result , PAPE finds that for hydrated sulphates of analogous formulae , the products of the specific heats and the equivalents are approximately equal ; and further , that with sulphates containing different quantities of water , the product of the specific heat and the equivalent increases with the quantity of water , in such a manner , that to an increase of each one equivalent there is a corresponding increase in the product .
15 .
In the preceding paragraphs I have collated , as far as I know them , the investigations on the specific heat of solid bodies , on the relations of this property to the atomic weight , and on the connexion with the chemical composition of a substance .
The views which have been expressed relative to the validity of DULONG and PETIT'S ?
and of NEUMANN'S laws , and also as to the question whether the elements enter into chemical compounds with the same specific heats which they have in the free state or with modified ones , have been various and often discordant .
In this respect it may be difficult to express an opinion which has not been already either stated or hinted at , or which at any rate cannot be naturally deduced from a view previously expressed .
The results to which my investigations .
on the specific heats of solid bodies have led me are the following:-Each solid substance , at a sufficient distance from its meltingpoint , has a specific heat , which may vary somewhat with physical conditions ( temperature , greater or less density , amorphous or crystalline conditions , &amp; c ) ; yet the variations are never so great as must be the case if a variation in the specific heat of a body is to the specific heat of crystallized borax and of ordinary phosphate of soda .
He has subsequently published the results of his experiments for the latter salt ( Ann. de Chim .
et de Phys. [ 3 ] vol. xxvii .
p. 253 ) , but I cannot find the number which he found for crystallized borax .
be held as a reason for explaining why the determinations of the specific heats of solid elements do not even approximately obey DULONG and PETIT'S law , nor those of solid compounds of analogous chemical constitution NEUMANN'S law .
Neither law is universally valid , although I have found that NEUMANN'S law applies in the case of many compounds of analogous atomic composition , to which , on account of their totally different chemical deportment , different formulas are assigned ; and even in cases in which these laws have hitherto been considered as essentially true , the divergences from them are material .
Each element has the same specific heat in its solid free state and in its solid compounds .
From the specific heats to be assigned to the elements , either directly from experimental determination , or indirectly by calculation on the basis of the law just stated , the specific heats of their compounds may be calculated .
I show the applicability of this by a great number of examples .
In reference to this calculation of the specific heats of solid bodies I may here make a remark .
The agreement between the results of calculation and experiment is often only approximate ; it is then natural to urge that the two ought really to agree more closely .
To that the question may be allowed : What means are there of even approximately predicting and calculating beforehand the specific heat of any inorganic or organic solid compound when nothing but its empirical formula is given ?
to which among the numbers 0.1 , 0-2 , 0-3 ... ... may it come nearest The cases in which differences exist between calculation and observation , enumerated in ? ?
103 to 110 , may be set against this uncertainty .
My proof of the propositions given above is based on determinations made by earlier inquirers , and on a not inconsiderable number of my own .
I first describe the method by which I worked , and then give the results which I have obtained by its means .
PART II.-DESCRIPTION OF A METHOD OF DETERMINING THE SPECIFIC HEAT OF SOLID BODIES .
16 .
I have worked by the method of mixture .
It is not necessary for me to discuss the advantages which this method has over that of the ice-calorimeter , at any rate in requiring smaller quantities ; nor , as compared with the method of cooling , need I discuss the uncertainties and differences in the results % for the same substance , which are incidental to the use of this method , and which REGNAULT has detailed* .
The method of mixtures has been raised by NEUMANN and by REGNAULT to a high degree of perfection .
Although by NEUMANN'S method it is possible to determine more accurately the temperature to which the body investigated is heated , REGNAULT'S method allows larger quantities to be used .
REGNAULT'S process gives the specific heats of such substances as can be investigated by it as accurately as can at all be expected in the determination of this property .
In the case of copper and steel , it is not merely possible to determine their specific heats by its means , but also to say whether and how far there is a difference in the first metal according as it has been heated or hanmtmered , and in the second , according as it is soft or hard .
It may be compared with a goniometer , which not only measures the angles of a crystal , but also the differences in the angle produced by heat ; or it may be compared to a method for determining the specific gravity of a body , by which hot only this property , but also its , changes with the temperature may be determined , But along with such methods , simpler ones , though perhaps less accurate , have also their value .
Which method is the most convenient or which ought to be used in a given case , depends on the question to be decided by the experiment , or on the extent to ' which the property in question is constant in the substance examined .
In regard to the relations of the specific heat of solid bodies to their atomic weight and to their composition , REGNAULT'S determinations have shown that both DULONG and PETIT'S and NEUMANN'S law are only approximate , and that even the accuracy in determining the specific heat which REGNAULT attempted , and obtained , could not show that these ' laws were quite accurate .
Although the description of REGNAULT'S mode of experimenting is so widely known , yet it cannot be said to have become the common property of physicists , or to have found an entrance into the laboratories of chemists , to whom the determination of the specific heat is interesting from its relation to the atomic weight .
Very few experiments have been made by this method other than the determinations of REGNAULT .
The method depends on the use of an apparatus which is tolerably complicated and takes up much room .
Each experiment requires a long time , and for its performance several persons are required .
REGNATJLT has usually worked with ' very considerable quantities of the solid substance , and in by far the majority of cases at temperatures ( usually up to 100 ?
) which many chemical preparations , whose specific heats it is important to know , do not bear .
In the sequel I will describe a process , for the performance of which the apparatus can be readily constructed , and for which one operator is sufficient ; by which , moreover , the determination of specific heat can be made with small quantities of the solid substance and at a moderate temperature .
But the method as I have used it has by no means the accuracy of that of REGNAULT .
In ? ?
1S I shall discuss the advantages for which some of the accuracy which characterizes REGNAULT'S method is sacrificed ; but I may here remark that the results obtained by the method which I have used are capable of increased accuracy , provided the experiments are executed on a larger scale and within greater ranges of temperature .
17 .
The principle which forms the basis of my method is as follows:-To determine the total increase of temperature produced when a glass containing the substance to be investigated , covered by a liquid which does not dissolve it , the whole previously warmed , is immersed in cold water ; to subtract from the total increase of temperature that due to the glass and the liquid in it , and to deduce from the difference , which is due to the solid substance , its specific heat .
If , in regard to gain or loss of heat , the glass , in so far as it comes in contact 84 with water , is equivalent to x parts of water , iff is the weight of the liquid in it , y its specific heat , m the weight of the solid substance , AM the weight of the water in a calorimeter , including the value in water of the immersed part of a thermometer and of the calorimeter , T the temperature to which the glass and its contents have been heated before immersion in water , and T ' the temperature to which the glass sinks when immersed in the water , while the temperature of the latter rises from t to t ' , then the specific heat ( sp. H. ) of the solid substance is MH ( t-t)-(x +y ) .
( T-T ' ) m(Trr ) In the sequel I shall discuss more specially the manner in which the individual magnitudes in this equation were determined : I will first give a description of the apparatus and method which I used* .
The glass vessel in which the substance is confined ( Plate XX .
a in fig. 1 ) is a tube of glass , the bottom of an ordinary test-tube .
In it fits , but not air-tight , a cork c , which is pressed between two small brass plates that are screwed to a wire b. The solid sub !
stance to be investigated , in the form of thin cylinders , or in small pieces the size of a pea , along with a liquid of known specific heat , which does not dissolve it , are placed in the tube in such a manner that the liquid covers the solid substance , and that there is a space between the liquid and the cork when it is inserted .
The glass , when the cork is fitted , may be suspended to the balance by the wire b. Three weighings ( 1 ) of the empty glass , ( 2 ) after introducing the solid substance , and ( 3 ) after introducing the liquid , give the weight of the solid substance ( m ) and of the liquid ( f ) .
The heating apparatus ( fig. 1 ) serves to raise the temperature of the glass with its contents .
The glass is dipped in a mercury-bath A near its upper edge , and retained by a holder E. The mercury-bath , which consists of a cylindrical glass vessel , is suspended by means of a triangle round the neck of the vessel in an oil-bath B , which stands on a tripod C , and can be heated by a spirit-lamp D. A thermometer dt , fixed to the holder F , is also immersed in the mercury-bath .
The flame of the spirit-lamp may be regulated so that the thermometer d indicates the same temperature for a long time : .
If it may be assumed that the contents of the glass a have also risen to this temperature , then the wire b being firmly held in the right-hand by its hook , and the clamp of the holder E in the left , the glass a is rapidly removed from the heating vessel to the calorimeter H ( fig. 2 ) .
This is almost the only part of the entire experiment which really requires much practice ; the transference of the glass a from the one vessel to the other must be effected in an instant , and none of the liquid in the glass must touch the cork .
The calorimeter H stands upon a support G ( fig. 2)* , on which there is an oval metal plate o. In this there are three depressions , in which fit the three feet of the calorimeter ( they are made of very thin hard brass wire ) .
The calorimeter is oval-shaped , and is made of the very thinnest brass plate .
In it a brass stirrer fits , made of two parallel plates of brass of the same thinness , which are joined below by thin wires , and provided with a thin wire ending in a little button i , which serves as handle .
The plates of the stirrer are perforated in such a manner that the glass a and a thermometer can be passed through them .
Fig. 4 shows more distinctly the construction of the stirrer , also the section of the calorimeter .
For the experiments , the calorimeter is always filled , as nearly as possible , with the same quantity of watert .
The stirrer is immersed , and a thermometerf dipping in the water gives its temperature , which is kept uniform by an upward and downward uniform motion of the stirrer .
When the tube a is brought into the water of the calorimeter , it is fastened$ in the clamp of the holder K , which is arranged like the pincettes used for blowpipe experiments , so that it stands on the bottom of the calorimeter , and then the stirrer is set to work .
This motion of the stirrer , and therewith of the water , must be moderate and uniform in all experiments ; this is of some importance for the uniformity and comparability of the experiments .
The temperature indicated by the thermometer f rises and soon attains its maximum , which continues for some time , and can be observed with certainty .
With this the experiment is concluded .
The tube a can be taken from the calorimeter , dried , and used for a new experiment .
The increase of temperature produced in the calorimeter by the tube a and its contents , would be incorrectly given if the warmth of the body of the operator , who moves the stirrer and observes the thermometer , acted on the calorimeter .
This is prevented by a glass screen ggg g , fig. 2 , which is fitted in the brackets h h , and above which the handle of the stirrer projects .
18 .
This process for determining the specific heat of solid bodies , the details of which are more minutely discussed in the sequel , has advantages over those hitherto prinIn making the experiment , the actual distance between the calorimeter and the heating apparatus must be greater than is indicated in the figure , but not so great that the glass a cannot , by a rapid motion of the arm , be transferred from the mercury-bath to the calorimeter .
t This is most conveniently effected by laying across it a bridge with a stem directed downwards ( fig. 3 ) , and adding water until it touches the point of the stem ; and the calorimeter , which now contains almost the requisite quantity of water , is placed on the balance , and the filling completed by means of the dropping-flask ( fig. 8 ) .
The construction of the latter is readily intelligible : it is held by the cork between two fingers , and by approaching the hand to the bottom of the flask water commences to drop .
When the flask is not in use the tube , which fits air-tight in the cork , is raised , so that it does not dip in the water , and thus the water is prevented from escaping .
+ Fig. 5 shows in a section the glass a , with its contents , and the thermometer f immersed in the water of the calorimeter .
86 PROlFESSOR KOPP ON THE SPECIFIC HEAT OF SOLID BODIES .
cipally used , which I will here mention .
The use of the mercury-bath makes it possible readily to produce , and maintain for any adequate length of time , any temperature desirable in such experiments .
The mercury-bath * shares with the air-bath the advantage that , to the substance heated in it ( in this case the tube and contents ) , nothing adheres when it is remo-ved which might influence the thermal effect in the calorimeter .
It has over the air-bath the advantage , that any body placed in it takes the temperature of the surrounding medium much more quickly through its entire mass .
The communication of heat to the solid substance is materially promoted by the circulation of the liquid between its particles ; the time necessary for the entire contents of the glass to become equally heated is a very short onet .
Moreover this very circulation of the liquid between the particles of the solid ensures a quicker and more uniform transmission of the heat of the contents of the glass to the water of the calorimeter ; the maximum temperature of this water is soon attained+ , although the transmission of the excess of temperature must take place through the sides of the glass .
The apparatus which I have just described is very simple .
It is readily constructed ; the chief point is to have two thermometers which have been compared with each other , one of them ( f ) graduated in tenths of a degree , while on the other ( d ) the tenth of a degree can be observed with certainty .
The apparatus does not require much space ; yet , while the experiment is being made , rapid changes in the temperature of the surrounding air must be avoided .
One observer only is required ( all the experiments described in the sequel have been made without assistance ) .
The experiments which I shall communicate prove that , by means of this apparatus , the specific heat of solid substances , even when only small quantities are taken ( in most cases I worked with only a few grammes ) , may be determined with an accuracy not much less than that attained with larger quantities in more complicated processes .
19 .
Yet , it is true , the accuracy of the results obtained by this process appears to be inferior to that attainable by the use of NEUMANN'S or of REGNAULT'S methods .
I have investigated many substances , determinations of which have also been made by these physicists .
I do not find that the numbers I have obtained deviate in one special direction from those which these physicists have found , which moreover sometimes differ considerably among themselves * ; but that the certainty of the results I have obtained is less , is shown by the fact that the results of different experiments with the same substance agree less closely with one another than do those 'of REGNAULT and of NEUMIANN .
That my determinations are less accurate is probably least due to the circumstance that I did not use certain corrections , for instance , that I did not allow for the loss of heat in the calorimeter between the time when the heated body was immersed and the maximuim temperature was attained .
I have endeavoured to diminish the uncertainty of the results from this source by having the temperature of the water in the calorimeter , before immersing the heated body , somewhat lower than that of the surrounding air .
I have endeavoured to ensure comparability in my results for different substances by always operating as much as possible under the same circumstances ; that is , I endeavoured always to produce in the water of the calorimeter the same excess of temperature over that of the surrounding air .
Without depreciating the interest and value of such corrections , I think that their application may be omitted if their practical importance is inconsiderable , and the increased difficulties which they necessitate proportionally large .
It must be considered , in reference to such corrections , how far the accuracy , which the results obtained by their means claim , is not more apparent than real $ .
And further , that these corrections , where the conditions for their application really exist , are not considerable ; while , where they exert a considerable influence on the result , they may be uncertain , because the suppositions made in their development X PAPE , in POGGENDORFF'S 'Annalen , vol. cxx .
p. 579 , discusses the probable causes of these differences .
t Another correction , which appears to me to be more important for the experiments in question , is , that the contents of the glass at the time at which the temperature of the water is at its maximum may be at a somewhat higher temperature .
This I have approximately taken into account .
Compare ? ? ? ?
23 &amp; 24 .
+ It is unnecessary to adduce examples where such corrections , proceeding from as comprehensive a basis as 88 are less applicable .
But more especially can such corrections be disregarded when , as in the case with my determinations , other circumstances diminish more materially the accuracy of the results to be obtained .
Such circumstances in my experiments are , that I worked on a small scale in every respect .
I could only heat the solid investigated together with a liquid to 500 , and in many cases not even to this .
In NEUIMANN'S and in REGNAULT'S experiments , on the contrary , the solid was usually heated to near 100 ?
, and the differenc'e in temperature , T-T ' ( compare ? ?
17 ) , obtained in the latter experiments was usually thrice as great as in mine .
In REGNAULT'S experiments ( in NEUMANTN'S the details are not given ) the quantity of substance taken was , on the average , twenty times as much , and the weight of water in the calorimeter about eighteen times as much as in mine * : hence in the latter experiments the unavoidable accidental errors of observation must be greater than in the former .
But there is a still more important circumstance which makes the accuracy to be hoped for from my experiments less than that to be expected from REGNAULT'S and NEUMIANN'S experiments .
In the latter methods the increase in the temperature ' of the water of the calorimeter is entirely , or is almost entirely produced by the solid under examination .
In my experiments , on the contrary , this increase is produced by the glass , the solid , and the liquid in the glass .
The thermal action due to the solid is only a part of the entire thermal action observed , and if from the latter that due to the liquid and to the glass is subtracted , all uncertainties in the assumptions as to the thermal action of the possible , lose their significance from necessary simplifications , and their practical impoitance becomes finally very slight .
The amount of correction is then to be pronounced as having no influence on the final result .
It is more important to take into account the following .
The trustworthiness of the specific heat to be assigned to any particular compound depends upon the certainty of the determination of the physical property , and upon the certainty of the knowledge of the composition of the body in question ; that is , in how far this compound corresponds to a given formula .
The greatest trouble which can be taken in that determination , the consideration of all sources of error which are possible in the physical experiment , the most complete exposition of the corrections which by developing conclusions from more or less certain assumptions may be formulated in one expression , and the most conscientious application of these corrections , -all this may be paralyzed by the circumstance that the composition of the body in question is not , as it were , the ideal , not corresponding accurately to the formula .
The partial substitution , if even to a very small extent , of one constituent by an isomorphous one , the attraction of water by a hygroscopic substance before and during the experiment , the , presence of some mother-liquor in a crystallized salt , the loss of some water in drying a hydrated substance , so that this has not exactly the composition corresponding to the formula , -all these sources of error , which can scarcely be taken into account , may easily exercise an influence on the final result , whose magnituLLde far exceeds that of certain corrections applied to the physical part of the determination .
It lies in the nature of the case that in such investigations , in some cases bodies of well-known , in other cases bodies of less well-known composition are taken .
I tried to be certain what substances could be considered as of definite composition and what of doubtful composition , especially where the relations between the specific heat and the atomic weight or chemical composition were under discussion , liquid and that of the glass are concentrated on the remainder , on the thermal action of the solid substance from which its specific heat is to be deduced .
The results obtained by my method are less accurate when the residue is only a small fraction of the total result from which it is deduced .
In individual cases , where this was unavoidable , I shall have to remark upon it .
It may be said in favour of my method that , for a number of solid substances , no other method yet attempted is applicable either at all or with more prospect of a successful result .
But this is less important than the proof furnished by my examination of very many substances , whose specific heat has been already determined by NEUMANN and by REGNAULT , that the specific heat of bodies may be determined by my method with an accuracy quite sufficient for many comparisons .
But there are cases in which it is even advantageous not to heat the solid alone , but in conjunction with a liquid , and to bring them together into the water of the calorimeter .
The chemical nature of the solid may necessitate this ; as , for example , when it readily alters on being heated in the air ( compare ? ?
34 in reference to amorphous boron ) ; its physical structure may also render it desirable , as for instance if the substance has a large surface as compared with its mass , or is so porous that the thermal action due to humectation , and first observed by POUILLET X , takes place .
REGNAULT has shown that this may be considerable t ; he states that for this reason the specific heat of some substances is found about --1 too great .
He appears to have estimated this thermal action by ascertaining the increase of temperature produced in the water of the calorimeter when the porous substance , whose temperature is that of the water and of the surrounding air , is dipped in it .
But this action is probably far more considerable if , while heated , it is immersed in the water , because it then contains less air confined on its surface and in its pores $ , and surface action can then act more intensely upon the liquid .
The influence of this source of error cannot be measured exactly .
It is unequal in different substances .
In platinum it is small ( REGNAULT found by his method that the specific heat of spongy platinum did not materially differ from that of massive pieces ) , while it may be con@ Ann. de Chim .
et de Phys. [ 2 ] vol. xx .
p. 141 .
t Ann. de Chim .
et de Phys. [ 3 ] vol. i. p. 133 .
REGNAULT preferred to immerse the heated porous substances , when they could be obtained in coherent pieces , directly in the water of the calorimeter .
If they were enclosed in thin tubes and immersed , the equalization of temperature proceeded too slowly .
REGNATJLT abstained from enclosing at the same time a sufficient quantity of water in the tube to promote the circulation , because in that case the thermal action of the solid was only a fraction of that of the water added , on which the entire source of error falls .
REGNAITLT found also ( ibid. p. 142 ) that in immersing anhydrous baryta , strontia , and lime in most carefully dehydrated oil of turpentine , there is such a thermal action that no useful result is to be obtained by his method for these oxides .
+ To the examples already known , which show what influence temperature exerts on the quantity of air absorbed in a porous body , REGNAUTLT has added a very instructive one ( Ann. de Chim .
et de Phys. [ 3 ] vol. lxiii .
p. 32 ) .
If amorphous boron , formed into disks by pressure in a steel mortar , was strongly cooled and then immersed in the water of the calorimeter ( at the mean temperature ) , so considerable a disengagement of absorbed air was produced , that REGNAULT was compelled to give up the determination of the specific heat by this method .
90 siderable for porous charcoal ( in fact POUILLET'S experiments make this probable ) .
This source of error is excluded in my method .
20 .
In order to appreciate the trustworthiness of the results arrived at by my mode of experiment , it is important to state with what amount of accuracy the data of observation and the ancillary magnitudes were determined .
I will give this statement in what now follows .
For observing the temperature of the water in the calorimeter I used thermometers made by GEISSLER of Bonn , which the kindness of Professor BUFF , Director of the Physical Cabinet in Giessen , placed at my disposal .
In these thermometers the tube consists of a fine glass thread drawn out at the lamp .
The bulb is cylindrical , very thin in the glass , and contains but little mercury .
On one ( b ) 1 ?
C. corresponds to a length of almost 5 millims. on the scale , and on the other ( r ) to almost 4-5 millims. Tenths of a degree can be read off directly on the scale , and it is easy to learn to estimate hundredths safely .
I have repeatedly compared these two thermometers , between 7 ?
and 24 ?
, with two normal thermometers of my own construction , which agree very well with each other , and on one of which a division corresponds to 004878 , and the other to 00 ?
4341 .
The differences of the indications between the GEISSLER'S thermometers and these could be considered as constant within those limits ; for the differences thus observed all the readings made with the GEISSLER'S thermometers had to be corrected to make them comparable with the indications of the normal thermometer .
The temperature of the mercury-bath was ascertained by means of one of these normal thermometers , and the indications of this thermometer immersed in the bath ( d in fig. 1 .
) corrected for the lower temperature of the mercury thread out of the bath ; this latter temperature was given with adequate approximation by the second thermometer , e. 21 .
The weight of the thin sheet-brass calorimeter , together with stirrer , was 11 145 grms.* Taking the specific heat of brass , according to REGNAULT , at 0*09391 , the calorimetric value in water of this mass of metal is 1 -046 grm. Considering that the calorimeter in the experiments was not quite filled with water , but about -*th remained empty , even after introducing the tube , I put the value in water at 0*872 .
In determining the calorimetric water value of the immersed parts of the thermometers r and b , the following experiments were made .
The weight.of water in the calorimeter , together with the reduced weight of the metal , was 30-87 grms. When the thermometer r heated to 33 ?
-86 was immersed , the temperature rose from 10 ?
-73 to 100 ?
85 ; the immersion of the thermometer b at a temperature of 37 ?
053 caused a rise from 100 ?
61 to 10 ?
.76 .
In both cases the temperature of the water was indicated by means @ At the beginning of these investigations .
During their progress the calorimeter was cleaned and dried with bibulous paper a countless number of times , so that its weight diminished by about 0-04 grm. in the course of the experiments .
In determining the weight of water used in each experiment , the weight which the calorimeter actually had at the time was taken as basis .
91 of the other thermometer , the reduced value of which might be neglected under these circumstances .
These experiments gave 0-16 as the reduced value of the thermometer r , and 0 17 as the reduced value of the thermometer b. The thermometers have very nearly the same dimensions .
Hence I put the reduced value of the calorimeter ( that is , of the part of the met , al concerned ) , of the stirrer , and of the immersed part of the thermometer at 1 04 grin .
Even if this determination is a few tenths out , it is scarcely appreciable as compared with the quantity of water in the calorimeter .
In all following experiments this was between 25 85 and 25-95 grms. All the subsequent determinations depend on fixing differences of weights and of temperatures .
The accuracy of the results depends on the precision with which both kinds of magnitudes are ascertained ; and it is useless to determine the weights to 1-,1-0 or nearer , if the differences in temperature cannot be determined more accurately than to -OOor 3-0 ?
I have weighed to centigrammes instead of to milligrammes , by which the time necessary for the weighings was much shortened , and their accuracy not materially lessened .
22 .
The reduced value x remained to be determined of the glasses ( cylindrical tubes of thin glass , see ? ?
17 ) , or , rather , of that part which was immersed in the water of the calorimeter , the quantity of which was always the same .
In the following , T is the temperature to which the glass in the mercury-bath was heated ( compare fig. 1 ) , M the quantity of water in the calorimeter + the reduced value in water of the other parts of the latter , which required to be taken into account , t the temperature of the water in the calorimeter when the glass was immersed ( fig. 2 ) , and r the temperature to which the water became heated , and which must be considered as that to which the glass cooled* .
We have then M(r t ) T-r In my experiments I used three glasses , which may be called 1 , 2 , and 3 .
To ascertain the reduced value of glass 1 , I made the following determinations : Temperature of Air 15'8 .
T. .
t. M. X. 000 grms. 78-54 17-23 15-72 26-98 0-664 74-38 17-16 15-78 26-97 0-651 75-51 17-14 15-72 26-92 0-655 76-06 17-15 15-73 26-945 0-649 77 32 17-22 .
15-74 26-96 0.664 Mean. .
0-657 subsequently made a second series of experiments to determine the reduced value for glass 1 , which gave the , following results : Temperature of the Air 19 ? ?
9-19 ? ?
8 .
T. .
t. M. X. ?
o0 grms. 78-50 21-32 19-93 26-99 0656 81-86 21-47 20-03 26-98 0-643 80-42 21-43 20 02 26-98 0-645 79-77 21-42 20-03 26-935 0-642 80-14 21-51 20-12 26 955 0-639 Mean. .
0645 The mean of these two means , 0-65t7 and 0-645 , gives as the reduced value in water of glass 1 , 0-651 grm. To obtain the water value for glass 2 , I made the following determinations : Temperature of the Air 12 ?
-0-12 ?
-5 .
T. r. t ... x. ?
oo grms. 75-87 13,53 12-43 26-94 0-475 77-05 13-46 12-31 2696 0-488 76-71 13 68 12-54 26-975 0-488 75-97 13-76 12-65 26 95 0-481 78-60 13-83 12-62 26-95 0-503 Mean. .
0487 The reduced value for glass 2 is hence = 0-487 grmn .
This glass broke before I made a second series of experiments to ascertain its reduced value .
I made two series of experiments to determine the reduced value of glass 3 .
The first gave the following results : Temperature of the Air 19 ?
-3-19 ?
-5 .
T. X. t. M. .
ogrns .
81-00 20-33 19.31 26-98 0-454 80-03 20-83 19-84 26-965 0-451 80-22 20-93 19-94 26-98 0-451 84-06 21-04 20-02 26-945 0-436 81-90 20-93 19-93 26-975 0-442 Mean. .
0447 The second series of experiments gave the following results : Temperature of the Air 19 ?
'9-19I-8 .
T. V. t ... 000 grms. 80-41 21-08 20-06 26-965 0-464 79-64 21-10 2009 26-965 0-465 79-98 21-12 20-12 26-96 0-458 80-22 21-12 20-12 26-985 0-457 79-53 21-10 20-12 26-965 0 452 80-52 21 13 20-14 26-96 0-450 Mean. .
0-458 The reduced value of glass 3= 0-453 grm. , the average of the mean numbers of both series of experiments .
23 .
In those experiments in which a glass containing a liquid and perhaps a solid substance is immersed , while warm , in the water of the calorimeter , it may be asked if , when the water has become heated to a certain maximum temperature , the contents of the glass have actually cooled to the same temperature .
In earlier experiments made by the method of mixture , it was at once assumed that the temperature assumed by the water of the calorimeter after immersing the solid was actually that also to which the immersed body sank .
NEUMANN has taken into account that the immersed body , when the water shows its maximum temperature , may have a somewhat higher temperature A. AVOGADRO has also taken this into accountt , and REGNAULT has also allowed for this circumstance in the case in which the mass , immersed in the water of the calorimeter , is a bad conductor of heat $ .
A correction for this fact is certainly inconsiderable and unnecessary if the immersed body conducts heat well , and the range of temperature through which it cools in the liquid is great .
This interval of temperature was in my experiments considerably smaller than in those of NEUMANN and of REGNAULT ; and as in my experiments the excess of heat of the contents of the glass had to pass through its sides to the water of the calorimeter , it might be doubted whether , when the temperature of the water was at its maximum , this temperature could be considered as that of the contents of the glass .
I have endeavoured to answer these questions experimentally .
A glass , such as was used for holding the solid investigated and a liquid , was filled with water , and a perforated cork fitted , by means of which the glass could be handled , and which permitted the introduction of a thermometer into the water within the glass .
The glass filled with water was warmed , and then placed in the calorimeter filled with water ; a thermometer A , passing through the cork , showed the temperature of the water in the glass ; e In the memoirs mentioned in ? ?
4 , P.PE has also discussed and applied the correction to be made for the above circumstance ( POGGENDORFP'S * Annalen , ' vol. cxx .
p. 341 ) .
t Ann. de Chim .
et de Phys. [ 2 ] vol. Iv .
p. 90 .
+ Ibid [ 2 ] vol. lxxiii .
p. 26 .
94 a second , B , showed that of the calorimeter water .
If the glass filled with the warmer water is immersed in the cold water , the following circumstances are observed* .
A sinks very rapidly , while B rises more slowly ; if B shows the maximum temperature for the water of the calorimeter ( this temperature being called t ' ) , A gives a higher temperature ( T ) for the contents of the glass .
B then slowly sinks and A follows it , while the difference between t ' and T ' always becomes smaller .
In the two following series of experiments I have endeavoured to determine by how much , under certain conditions , the temperature T ' of the water in the glass exceeds the maximum temperature t ' of the water in the calorimeter when this maximum temperature as such is observed .
I obtained the following results : the temperature of the air in the experiments was 13 ? ?
2-13 ? ?
5 .
Experiments with Glass 1 .
Experiments with Glass 2 .
T ' .
t ' .
Difference .
T. t ' .
Difference .
15-51 18-13 0-38 15-71 15-50 0-21 14-96 14-72 0*24 15-96 15-65 0 31 16 11 15-94 0.17 15-16 14-91 0-25 15-56 15-36 0-20 14-76 14-47 0-29 14-24 14-05 0419 14-66 14-33 0-33 15-96 15-64 0-32 15-56 15-24 0.32 A closer agreement in the numbers expressing the difference between T ' and t ' is difficult to attain , since a certain time is necessary to observe the occurrence of the maximum temperature , and during the time in which the thermometer B remains constant , the thermometer A still sinks ; according to the moment at which the maximum temperature is considered to be established , this difference may be obtained different , and the smaller the later the observation is made .
Moreover the magnitude of this difference between T ' and t ' depends on the difference between t ' and the temperature of the air .
I have always endeavoured to work under the same circumstances , and especially to arrange the experiments so that the maximum temperature of the water in the calorimeter did not exceed by more than 2 ?
the temperature of the airt .
For these experiments and the apparatus which I used , I assumed , on the basis of the preceding experiments , that if the water of the calorimeter had assumed its maximum temperature t , the contents of the glass were 0 ?
-3 higher ; that is , I put throughout T ' , the temperature to which the contents of the glass immersed in the calorimeter had fallen , =t'+0 ?
0 3 .
24 .
It is a matter of course that , in introducing this correction for obtaining the tem* In these experiments , in order to ensure uniformity in the temperature of the water , the stirrer was kept in continual motion , and the same process followed as in ascertaining the specific heat .
fA preliminary experiment shows how cool the water in the calorimeter ought to be .
Water which is somewhat cooler than the surrounding air , may be kept in stock for such experiments by placing it in a cylindrical flask covered externally with filtering paper , and standing in a dish of water , so that the paper is alwaysa moist .
To warm the water in the calorimeter , it was merely necessary , with apparatus of the dimensions I used , to lay the hand on it for a short time .
perature of the contents of the glass at the time the maximum temperature has been attained in the calorimeter , it is unnecessary to give the indications of T ' in hundredths of a degree ; and since the temperature T , to which the glass with its contents was heated in the mercury-bath , only serves to deduce the difference T-T ' , it is unimportant in giving this temperature to do so in hundredths of a degree .
The accuracy of the determinations of specific heat , in so far as it depends on determinations of temperature , is limited by the accuracy with which the difference of T-T ' and t'-t are determined ( where t is the original temperature of the water in the calorimeter , and the other letters have the meanings previously assigned to them ) .
To have one-of these differences very accurately , while the other is much less accurately determined , avails nothing for the accuracy of the final results .
It is at once seen that in my experiments , and especially in those of NEUMANN and REGNAULT , the hundredths of a degree have a greater significance for the small difference t'-t , than the tenths of a degree for the great difference T T ' .
The correction for educing the value of T ' , which I have just discussed , is of course more important the smaller the difference T-T ' ; for most of my experiments in which this difference is about 30 ?
, the significance of this correction is inconsiderable , if the contents of the glass be a good conductor .
I give a few numbers .
The experiments given in ? ?
25 on the specific heat of mercury , which , by using this correction , give it at 0 0335 in the mean , give it =0-0331 if this correction is neglected , that is , T ' made=t ' .
The fourth series of experiments , given in ? ?
27 , for determining the specific heat of coaltar naphtha A , give it at 0 425 when this correction is made , and at 0-420 when it is omitted .
The first series of experiments in ? ?
33 , for determining the specific heat of sulphur , give it at 04159 when this correction is used , and at 04152 when it is neglected .
Whether in all such cases T ' is put =t ' , or= t'`+003 , is of inconsiderable importance .
The correction in question is inadequate if the substance in the glass is a bad conductor ; for example , when the solid in the glass is a pulverulent or porous mass , in which the moistening liquid stagnates ( compare ? ?
18 ) .
That , under such circumstances , the numbers obtained for the specific heat are found somewhat too small must be remembered in ? ?
41 in the case of chromium , and in ? ?
52 in the case of chloride of chromium .
Too small numbers are also obtained , if in the experiments the maximum temperature of the cooling water exceeds that of the air by much more than 2 ? .
Such experiments are not comparable with the others , for example , with those made for the purpose of ascertaining the ancillary magnitudes occurring in the calculation of the results ; for them this correction is inadequate , and the loss of heat which the contents of the calorimeter experiences between the time which elapsed between immersing the glass and the establishment of the maximum temperature is too great .
By individual examples in ? ?
25 in the case of water , in ? ?
39 in the case of copper , and ? ?
41 in the case of iron , I shall call to mind how this source of error may give somewhat too small numbers for the specific heat ; but I have always tried to avoid this error , since I saw its importance in my first preliminary experiments .
96 TlIE SPECIFIC HEAT OF SOLID BODIES .
25 .
I first attempted to test my method by some experiments in which water or mercury was placed in the calorimeter .
For the specific heats of these liquids the following numbers were obtained , calculated by the formula HM ( t')-x ( T-T ' ) sp. H spH= f ( T-T ) in which the signification off is manifest from what follows , that of the other letters from what has been given before .
In the experiments in which a readily vaporizable liquid was contained in the glass , such as water , or coal-tar naphtha , a sensible formation of vapour took place , although the temperature did not exceed 500 .
If the glass containing the liquid was heated in the mercury-bath ( compare fig. 7 ) , vapour was formed in the empty space below the cork which served as stopper ; if the glass was then brought into the water of the calorimeter , this vapour condensed and settled partially on the stopper .
The stopper did not act materially on the water of the calorimeter ( see fig. 5 ) .
The quantity of liquid in the glass which acted directly on the water of the calorimeter , decreased somewhat in each experiment ; but this decrease is very inconsiderable .
In the following experimentsf denotes first the weight of the liquid in the glass at the commencement of the experiment , and at last its weight at the end of the experiments , that is , after subtracting the liquid which had vaporized and condensed on the stopper .
After the end of the experiment the stopper was dried to remove the liquid , and by another weighing of the glass , together with its contents and stopper , the weight of the liquid still contained in the glass was obtained .
The decrease of weight of the liquid in the glass was always found to be inconsiderable , and might without any harm have been neglected ; for the last experiment of a series I have always taken the diminished weight of the liquid into account , but for those between the first and the .
last I have neglected the diminution of the weight of the liquid in the glass .
What I have here said explains a remark of frequent subsequent use , " after drying the stopper .
" In reference to the influence of the formation of vapour on the accuracy of the results obtained for the specific heat of the individual substances , compare ? ?
38 .
Two series of experiments in which water was contained in the glass , gave the following results for the specific heat of this liquid : Experiments with Glass 1 .
Temperature of the Air 19 ?
*0 .
, T T ' .
t ' .
t. M. f. .
sp. H. 0 grms. grms. grm. 45-2 209 20-62 16*83 26-945 3-43 0-651 1P035 46-6 21`2 20*92 17-03 26-935 , , , 1013 47-4 21'3 20-96 17-03 26-965 3.42* , 0.997 " After drying the stopper .
97 Experiments with Glass 3 .
Temperature of the Air 19 ?
40 .
T. T ' .
t ' .
t. M. f. X. sp. H. oo 0o o grms. grms. grm. 46-8 21-1 20*76 17-03 26-95 3-445 0-453 1-004 46-8 21-1 20-83 17-12 26-985 , , , 0.999 47-0 21-2 20-93 17-22 26-935 3-435* , 0-996 The value found for the specific heat of the contents of the glass comes very near the number 1 , assumed for the specific heat of watert .
Determinations in which mercury was contained in the glass gave the following results for the specific heat of the contents of the glass .
Experiments with Glass 1 .
Temperature of the Air 13 ?
-8-14 ?
-4 .
T. T. t ' .
t. M. f. .
sp. H. 5 ?
0 0 ~8 10 0 3 ?
Egrms grmis , grm. 51 1 16-8 16-50 13-41 26-945 53-015 0-651 0-0335 48-5 16-8 16-48 13-64 26-95 , , , , 00333 45-2 16-5 16-20 13-63 26-965 , , , , 00333 Experiments with Glass 2 .
Temperature of the Air 13 ?
08-140 ?
4 .
T. T ' .
t ' .
t. M. f. w. sp. H. oooo grms. grms , grm. 50.0 17-1 16-79 13-74 26-935 60-015 0-487 0-0335 45-6 16-7 16-41 13-72 26-935 , , , , 00337 The mean of these five determinations gives 0-0335 for the specific heat of mercury , in accordance with the results found by other observers for this metal ( 0-0330 between 0 ?
and 100 ?
, DULONG and PETIT ; 0-0333 , REGNAULT ) .
26 .
For the liquid which is to be placed in the glass along with the substance whose specific heat is to be investigated , I could in many cases use water ; but many substances , the * After drying the stopper .
t In ? ?
24 it was mentioned that the numbers obtained for the specific heat of the contents of the glass are somewhat too small , if the maximum temperature of the water in the calorimeter , t ' , exceeds the temperature of the air by much more than 20 .
As an example I give the following determinations , in which the glass used contained water .
Experiments with Glass 1 .
Temperature of the Air 13 ? ?
5-13 ? ?
8 .
T. T ' .
t ' .
t. M. f. .
sp. H. o0o o0 grms. gris .
grm. 46-5 18-1 17-81 13-64 26-94 3-40 0-651 0-976 43-9 16.7 16-38 12-33 26-955 , , 0-989 Experiments with Glass 2 .
Temperature of the Air 130-5-130 ?
8 .
T. T. t ' .
t. M. f. x. sp. H. oooo grms , grs , .
g ?
rm .
4941 18.3 18-03 13-37 26-94 3-66 0-487 0-981 47-6 18-3 18-04 13-66 26-99 , , , , 01969 47-0 17-5 17-22 12*73 26-97 3.65* , , 0.991 determination of which is important , dissolve in , water , and hence I had to use a different liquid .
Coal-tar naphtha has the advantage that it is a mobile liquid , does not dissolve most salts , and does not resinify in contact with the air ; but besides the disagreeable odour , with continuous working , respiring air charged with its vapour appears to act injuriously on the organs of the voice .
As compared with water , coal-tar naphtha has the disadvantage , that its specific heat must be specially determined , and any possible uncertainty in this is transferred to the determination of the specific heat of the solid substance ; but the thermal action of a given volume of naphtha is only about that of the same volume of water* ; and in experiments in which the thermal action of a solid substance is 'determined , along with that of the necessary quantity of liquid which is contained with that substance in a glass , the thermal action due to the solid is a larger fraction of the total if coal-tar naphtha is used than if water is the liquid , which is a favourable circumstance in the accurate determination of specific heat .
As it was more especially important for me to obtain comparability in the results for specific heat , I have , for a great many substances which are insoluble in water , and for whose investigation water might have been used , also employed coal-tar naphtha .
Water was used for a few substances which are soluble in coal-tar naphtha ( sulphur , phosphorus , sesquichloride of carbon , for instance ) .
Several substances I determined both with water and with naphtha ; the results thus obtained agree satisfactorily .
To the question as to whether any possible change in the specific heat of naphtha with the temperature can be urged against the use of this liquid , I shall return in ? ?
29 .
27 .
The coal-tar naphtha A which I principally used in the subsequent experiments was prepared from the commercial mixture of hydrocarbons C , , H2-6 , by purifying it by means of sulphuric acid , treating the portion which distilled between 1050 and 120 ?
with chloride of calcium for six days , then again rectifying it , and collecting separately that which passed between 105 ?
and 120 ? .
This liquid had the specific gravity 0 869 at 15 ?
; in determining its specific heat I made four series of experiments , two at first when I was engaged on experiments in which I used this naphtha , and two towards the end .
I.-Experiments with Glass 1 .
Temperature of the Air 120 ?
1-12 ?
-9 .
T. T ' .
t'X t. M. f. x. sp. H o0 00 grms , grms , grm. 46-1 13'8 13-51 11-24 26-99 2-875 0-651 0-433 48-6 14-0 13-71 11-24 26-945 2-875 , ,t 0-443 45-5 14-1 13-83 11-59 26-97 2-975 4 0-439 45-3 14-3 14-01 11-80 26-94 2-970 f , 0-428 Mean .
0-436 II.--Experiments with Glass 2 .
Temperature of the Air 12 ?
'1-12 ?
*7 .
M. grms. 26-955 26-93 26-95 26-95 fx .
sp. H. grins , grm. 3'28 0-487 0-438 3'48 * 0-427 , , , , 0-427 3-475 , , 0435 Mean ... 0-432 T. 51-4 51'5 515 ' 51-0 III.-Experiments with Glass 1 .
T ' .
t ' .
t. 18-6 18-32 16-02 18-4 18-06 15-73 18-4 18-14 1581 18-5 18-22 15'93 IV.-Experiments with Glass T. T ' .
tt .
t. 517 171 18-43 1622 50-7 18-6 18-32 16-14 50'7 18-6 18-27 16-13 50-2 18-6 18-26 16-14 Temperature of the Air 16 ?
'7 .
M. f. .
sp. H. grms. grms. grm. 26-98 2-895 0-651 0-429 26-97 , , 0-431 26-985 , , , 0431 26-96 2-88t , 0-434 Mean ... 0431 Temperature of the Air 16 ?
'7 .
M. grms. 26-935 26-935 26-95 26-93 f. grms. 3'195 3.18t Mean 3-18 t Mean X. sp. H. grin .
0-453 0-423 , 0-431 0-421 0-426 ... 0'425 The average of the means of these four series of experiments , 0-436 , 0'432 , 0-431 , 0*425 , gives 0'431 as the specific heat of the coal-tar naphtha A between 14 ?
and 52 ?
; this value is taken in calculating the experiments in the following section .
28 .
If it were only a question as to the determination of the specific heat of this naphtha , the method described in the preceding might be advantageously replaced by another .
For by this method the specific heat of the liquid must be found somewhat too great , owing to the fact that in the empty space in the glass under the stopper a distinct quantity of vapour is formed , which condenses when the glass is dipped in the water of the calorimeter ( compare ? ?
25 ) .
Direct experiments , in which this formation of vapour was almost entirely avoided , have shown that the method used for the previous determinations , that is , the use of glasses for heating the liquid in which a* After adding some naphtha .
t After drying the stopper .
relatively considerable space above the liquid remains empty , gives the specific heat of readily vaporizable liquids somewhat too high , but that at the same time this influence of the formation and condensation of vapour is very small in the conditions under which I worked*.-The number 0-431 obtained in the previous determinations expresses the thermal action due to the cooling of 1 grm. naphtha A through 1 ?
in my experiments , which thermal action depends to by much the greatest extent on the specific heat of this liquid , and only to a very small extent on the condensation of the previously formed vapour .
In calculating the experiments communicated in the third section , that number is taken as the expression for the thermal action of naphtha , which is put as proportional to the weight of the latter .
This is , strictly speaking , not accurate , in so far as the thermal action arising from condensation of vapour only depends on the magnitude of the empty space and the temperature , and not on the quantity of naphtha in the glass .
But the small possible inaccuracy due to this cause in my experiments is not to be compared with other uncertainties .
The manner in which I have taken into account the naphtha contained in the glass corresponds most accurately to the actual conditions of the experiment , when this thermal action is most considerable ( only naphtha in the glass ) ; and if my mode of calculation less satisfies these conditions ( less naphtha in the glass ) , the entire amount is less considerable , and the influence of that which might be missed in that calculation , a vanishing quantity .
29 .
My experiments have been made at very different temperatures .
The temperature of the air was often something under 10 ?
, sometimes above 20 ?
, These numbers represent the limits to which the liquid in the glass , together with the solid substance cooled in the calorimeter .
In most experiments I heated the glass with its contents to about 50 ?
, in some cases not so high .
Now , for the various intervals of temperature within which the liquid in the glass cooled , can its specific heat be assumed to be always the same !
For water this may be done , and for coal-tar naphtha I did not calorimeter ( comp .
fig. 6 ) , was = 0-688 grm. A series of experiments in which this glass was used to determine the specific heat of the naphtha A gave the following results : Temperature of the Air 15'.5-15-06 .
T. T ' .
t ' .
t. M. .
x. sp. H. ooo grms. grms , grm. 52-5 17'8 17'53 14-93 26-945 3-205 0-688 0.415 49-6 17-4 17-13 14'73 26-955 , , , , 0412 50'9 17-6 17-29 14'83 26-96 , , , 0-407 50-5 17-6 17-26 14-83 26-975 , , , , 0407 51-6 17*7 17-38 14-84 26-985 , , , 0-416 50-9 17-8 17-47 15-03 26-94 , , , 0-405 Mean ... 0-410 i This is seen from the experiments on water communicated in ? ?
25 , and from the subsequent determinations in the next section , in which water was contained in the glass along with the solid substance .
doubt it while engaged in my experiments .
I first , when they were finished , became acquainted with REGNAULT'S * investigations on the specific heat of liquids at various temperatures ; according to these experiments the specific heat of some liquids considerably increases with the temperature .
I have not directly investigated coal-tar naphtha in this respect , but it is probable that the specific heat of this mixture of hydrocarbons C H2 -6 , alters but little with the temperature , and it is certain that this change is without influence on the accuracy of my determinations of the specific heats of solid substances .
REGNAULT'S experiments t , made by the method of cooling , show no change for benzole , 61 H6 , between 20 ?
and 50 , while there is a distinct change in the case of alcohol .
For pure benzole $I found the specific heat by the method of mixture to be 0'450 between 46 ?
and 19 ?
; REGNAULT ?
found it between 71 ?
and 21 ?
to be 0-436 .
These numbers , obtained with different preparations , are not indeed comparable for a decision of the question just discussed , but they render improbable a considerable increase in the specific heat of benzole with the temperature .
What I more especially lay weight upon is this : the specific heats of solids which I have determined at various temperatures , by their agreement with the numbers previously found by others , do not indicate any influence of a change of specific heat of naphtha with the temperature .
30 .
My stock of the naphtha , discussed in ? ?
27 , was used before I had investigated all the solid substances , for which a determination of the specific heat appeared necessary .
Another quantity of the same coal-tar naphtha was subjected to the same treatment as indicated there , and the portion passing over between 105 ?
and 120( ) used for the remainder of the experiments .
To ascertain the specific heat of this naphtha B , I made the four following series of experiments : I.-Experiments with Glass 1 .
Temperature of the Air 18 ?
'1-18 ?
'3 .
T. T ' .
t t ' .
t. f. x. sp. H. oo~~ ~~o oO grms. grms. grm. 51i5 19'6 9'33 17-22 26-96 2-70 0'651 0-419 52-7 19'9 19-64 17-49 26-95 , , , , 0-413 50'5 19'8 19-54 17-51 26-99 , , , 0420 49-9 20-0 19-73 17'75 26'995 2'695 1 , , 0-422 Mean ... 0418 II.-Experiments with T. T ' .
t ' .
51-4 19-7 19-36 51 5 19 9 1963 49-1 19-9 19-61 50.5 20-1 19-82 Glass 3 .
t. 17-32 17-56 17-73 17-86 Temperature of the Air 18 ?
'1-18 ?
'3 .
M. f. x. sp. H. grms. grms. grm. 26-94 3-085 0-453 0'415 26-965 , , , , 0426 26-955 , , , , 0-416 26-98 3-08 * , 0-418 Mean ... 0-419 III.-Experiments with Glass 1 .
T. T ' .
t ' .
t. 20'0 19-8 20-0 20-2 20-2 20-2 19-49 19-73 19-92 19-86 19-86 17-27 17-64 1782 17-76 17-85 IV.-Experiments with Glass 3 .
T ' .
t ' .
t. 19'7 19-43 17-33 20-1 19-77 17-66 20-2 19-87 17-65 20-1 19-83 17-82 20-2 19-93 17-82 remperature of the Air 17 ?
'8-18 ?
*3 .
M. f. x. sp. H. gris .
grins .
grm. 26-97 2-80 0-651 0-427 26'96 , , 0-425 26-98 , , , , 0420 26-99 , , , , 0418 26-95 2-785* , 0-410 Mean ... 0420 Temperature of the Air 18 ?
'4 .
M. f. x. sp. H. grms. grms. grn. 26-96 3-31 0-453 0-424 26-99 , , , , 0416 2696 , , , , 0423 26-95 , , , , 0409 26-97 3-29 * , , 0417 Mean ... 0418 The average of the means of these four series of experiments , 0-418 , 0-419 , 0-420 , 0-418 , gives 0-419 for the specific heat of coal-tar naphtha B between 20 ?
and 50 ? .
In the preceding method of experiment , whether water or naphtha of the kind described is contained in the vessel , a temperature much higher than 50 ?
cannot be employed ; for otherwise the quantity of liquid evaporating and condensing on the stopper becomes far too considerable .
Perhaps with hydrocarbons of higher boilingpoints higher temperatures might be ventured upon : I have no experiments on this subject .
PART III.-DETERMINATION OF THE SPECIFIC HEAT OF INDIVIDUAL SOLID SUBSTANCES .
31 .
By the method whose principle and mode of execution have been discussed in the .
preceding , I have determined the specific heat of a large number of solid substances .
I should have liked to include a still larger number of bodies in my investigations ; but a limit was put by the straining of the eyes from constant reading of finely divided scales , and by the injurious action which the long-continued working with coal-tar naphtha produces .
My crystallographic collection furnished me with much material for investigating the specific heat of both naturally occurring and artificially prepared substances , but for much more I have to thank others .
By far the greater part of the chemical preparations investigated I obtained from the Laboratory of the University of Giessen , through the kindness of the Director , Professor WILL , and of the assistants , Professor ENGELBACH , to whom my thanks are especially due , Drs. K6RNER and DEHN .
Professor WOHLER , of Gottingen , placed a number of chemical preparations at my disposal .
Professor BUNSEN , of Heidelberg , has helped me to the investigation of some rubidium-compounds .
Platinum and iridium I have been furnished with by M. HERAUS , the proprietor of the well-known platinum-manufactory in Hanau .
I have had a very large number of minerals from the mineral collection of the University of Giessen , through the kindness of the Director , Professor KNOP ; and to obtain the necessary quantity of dioptase , Professors BLUM of Heidelberg , and DUNKER of Marburg , have contributed .
32 .
The signification of the letters in the statement of the following experiments and their calculation is clear from ? ?
17 ; in reference to the value of the numbers for M , compare ? ?
21 , for x ? ?
22 , for T ' ? ?
23 , for y ? ?
27 and ? ?
30 .
It would require too much space always to give the comparison of my results with those of other observers .
I can only do this in individual cases where there are considerable differences and their discussion is of importance .
For other substances , where there are recent observations by trustworthy observers , the Tables in ? ?
82 to ? ?
89 give data for comparison .
33 .
Sulphur : pieces of transparent ( rhombic ) crystals from Girgenti .
I made three series of experiments with this substance .
I.---Experiments with Water .
Glass 1 .
Temperature of the Air 130 ?
2 .
T. T ' .
t ' .
t. M. w. X. y. .
sp. IE .
~~o oo Qo grms. grms. grm. grm. 45-8 15-5 15-24 11-74 26-95 4-16 1-765 1-000 0-651 0-168 46-0 16-2 15-93 12-52 26-935 , , , , 0.160 45-2 16-0 15-73 12-42 26-945 , , , , , 0-153 45-8 16-4 16-05 12-74 26-96 , , 175 , , , 0-153 Mean 0 .
0159 * After drying the stopper : compare ? ?
25 .
104 II.--Experiments with Water .
Glass 2 .
Temperature of the Air 130'2 .
T. T ' .
t ' .
t. M. m. f. y. x. sp. .
g rms .
grns , grms. grm. 45-8 16-4 16-07 12-36 26-96 4-815 2-09 1-000 0-487 0-171 47-3 16'6 16-33 12-46 26-95 , , , , , , , , 0170 44-1 16-5 16-15 12-74 26-925 , , , , , , , , 0156 45-1 16-6 16-28 12-77 26'96 , 2-07* , 0159 Mean ... 0164 Both these series of determinations are from the time when I first worked at this subject .
Towards the end , when I had acquired tolerable readiness , I made a third series , which agreed very closely with the results previously obtained .
III.-Experiments with Water .
Glass 3 .
Temperature of the Air 17 ?
'2 .
T. T ' .
t ' .
t. .
m. f. y. x. sp. H. ooo grms. grms. grms. grm. 43-7 19-1 18-83 15-79 26-99 4-92 2-065 1-000 0-453 0-166 43-5 19-1 18-84 15-84 26-97 , , , , , , , , 0162 43-3 19-2 18-92 15-92 26-94 , , , , , , , , 0170 43-1 19-2 18-87 15-93 26-98 , 2-05 , , , , 0-166 Mean ... 0166 Taking the mean of the means obtained in the three series of experiments , 0-159 , 0-164 , 0-166 , we obtain 04163 as the specific heat of rhombic sulphur between 17 ?
and 45 ? .
By the method of cooling , DULONG and PETIT found the specific heat of sulphur at the mean temperature to be 0-188 ; NEUMANN found 0'209 by the method of mixture ; for sulphur which had been purified by distillation , fused and cast in rolls , REGNAULT found f the specific heat between 14 ?
and 98 ?
to be 0-2026 .
In these experiments a development of heat depending on a change from amorphous sulphur into rhombic-crystallized appears to have cooperated , and to have caused the circumstance observed by REGNAULT , that after immersing the heated sulphur in the water of the calorimeter , the maximum temperature was only set up after an unusually long time .
Sulphur which has solidified after being melted , usually contains an admixture of amorphous sulphur , the greater the more the melting-point has been exceeded , which at the ordinary temperature passes slowly , at 100 ?
more rapidly , into crystallized , accompanied by disengagement of heat .
The transformation of the sulphur set up by the heating , and continued in the water of the calorimeter , brought about this slow appearance of the maximum temperature , and made the specific heat appear too great ; for REGNAULT'S subsequent determinations $ , also made between 97 ?
and 99 ?
and the mean temperature , gave it considerably less : 0-1844 for freshly melted sulphur ( in which superfusion had been avoided ?
) ; 0*1803 for sulphur which had been melted two months ; 041764 for what had been melted two years ( and which had then given 0'2026 ) ; 0-1796 for sulphur of natural occurrence .
The difference between the latter result and my own doubtless depends , partially at least , on the fact that REGNAULT'S determination was made between 14 ?
and 99 ?
( the latter of which temperatures is very near the meltingpoint of rhombic sulphur ) ; mine was made between 17 ?
and 45 ?
' .
Tellurium : crystalline pieces t. Experiments with Naphtha A. Glass 3 .
Teinperature of the Air 18 ?
'6-19 ?
'1* T. ' T. t ' .
t. M ... y. .
sp. H. ooo grms. g rns .
grn.r grm. 51-8 20'4 20-07 17-96 26'93 10-80 1-93 0-431 0-453 0-0486 51'3 20-3 20-02 17-93 26-98 , , , , , , , 00495 51-5 20-7 20-36 18-33 26-93 , , , 0-0454 51-0 20*7 20'43 18-43 26'955 , 191$ , , 00466 Mean ... 00475 34 .
Boron.-I have made some experiments with this substance , which have some interest for the question whether this body has essentially different specific heats in its different modifications ; but the results are not very trustworthy , owing to the spongy nature of the amorphous boron and the doubtful purity of the crystallized variety .
The amorphous Boron ?
which I investigated was pressed in small bars , and had stood several days in vacuo over sulphuric acid .
Experiments with Naphtha A. Glass 1 .
Temperature of the Air 17 ?
0--17 ?
02 .
T. T ' .
t ' .
t. M. m. y. x. sp. H. 0o o grms. grm. grms. grm. 49-0 18-7 18-73 16-36 26'955 1-52 2-515 0-431 0-651 0-246 48-1 18'6 18-55 16-23 26-965 , , 0-254 48'0 18'6 18-64 16-33 26-95 , , , , , , 0-252 47-9 18-7 18-72 16-42 26-95 , , 2-49 , , 0262 Mean ... 0-254 Even if the results of the individual experiments agree tolerably with each other they are not very trustworthy ; for the quantity of boron ( only 1 grm. ) is very small , and the amount of heat due to the boron is a very small part of the total ( comp . ? ?
19 ) .
Yet I do not consider the result of the above series of experiments ( that between 180 and 48 ?
the specific heat of amorphous boron is about 0-254 ) as being very far from the truth .
There are no considerable accidental errors of observation in these experiments , to judge from their agreement with one another .
Of the constants for calculating the experiments , x and y must be taken into account in regard to any possible uncertainty .
It has been assumed that xz-0615 and y-=0431 ; if we took x-=063 and y=0*41 , the specific heat as the mean of four experiments would be =030 ; if x were 0'67 and y 0'45 , the specific heat would be 0*21 .
But from what has been communicated in ? ?
22 and ? ?
27 in reference to the determination of x and y , it cannot be assumed that any possible uncertainty in reference to these values can reach either of the above limits .
It can be assumed with the greater certainty that the specific heat of amorphous boron is between 0'2 and 03 and nearly 0'25 , because x and y could not simultaneously both be found too great or too small ( if x had been too small y would have been too great , and vice versed ) .
Crystallized Boron * .
Experiments with Naphtha A. Glass 3 .
Temperature of the Air 18 ?
-9-180'7 .
T. T ' .
t ' .
t. M. m. f. y. x. sp. H. ooo0 grms. grms. grm. grm. 50-9 20-8 20-52 18-53 26*94 2-82 1-53 0-431 0'453 0-237 51'3 20-8 20-52 18-52 26'975 , , , , , , 0233 51-5 20-8 20-53 18-53 26-985 , , , 0-229 51,4 20'8 20-46 18-43 26-99 , , 152t , , , , 0'222 Mean. .
0-230 Hence the specific heat of the crystallized ( adamantine ) boron investigated is 0'230 between 21 ?
and 51 ?
; it is pretty near that found for amorphous boron , 0'254 .
REGNAULT found : ( between 98 ?
and 100 ?
and the mean temperature ) 0-225 for a specimen of crystallized boron prepared by ROUSSEAU ; 0'257 for one prepared by DEBRAY ; 0-262 for one obtained from DEVILLE ; and 0'235 for a specimen of graphitic boron prepared by DEBRAY .
The specific heat of amorphous boron could not be determined by REGNAULT'S method , because , when heated to 100 ?
in air , it partially oxidizes into boracic acid with disengagement of heat ( three experiments , in which the quantity of boracic acid formed was determined , and its specific heat , but not the thermal action due to the formation of hydrated boracic acid in immersion in water allowed for , gave respectively 0'405 , 0'348 , and 0-360 , which numbers REGNAULT does not consider as even approximately representing the specific heat of amorphous boron ) , and when greatly cooled disengages a quantity of air when immersed in warmer water , which renders the results uncertain .
Phosphorus.-I have only made a few determ tions with ordintiary yellow phosphorus , which was cast in sticks .
Experiments with Water .
Glass 1 .
Temperature of the Air 10 ?
'9 .
T. T ' .
t ' .
t. M. m. f. y. x. sp. H. oo ogrms .
grms. grms. grm. 38-8 13-5 13-20 10-05 26-95 3-075 2-065 1-000 0-651 0-208 33-8 12-9 12-62 10-03 26-97 , , , , , 0204 35-5 13-2 12-91 10-17 26-93 , , 206* , , , , 0.195 Mean ... 0-202 The specific heat of yellow phosphorus , as deduced from these determinations , is somewhat greater than that found by REGNAULT , doubtless because in my experiments the upper limit of temperature , T ' , was nearer the melting-point of phosphorus , 44 ? .
Compare ? ?
82 .
Antimony.-Purified by LIEBIG'S method ; crystalline pieces .
I.-Experiments with Naphtha A. Glass 2 .
Temperature of the Air 14 ?
'7 .
T ' .
16-0 15'9 15'8 t ' .
15-65 15-64 15-53 t. 13-42 13-54 13-52 M. 26-945 26-98 26-96 m. grms. 12-245,5 , , f. y. x. sp. H. grm. grin .
1-925 0-431 0-487 0-0539 , , , , , , 0-0520 1.91* , , , , 0-0496 Mean ... 0-0518 II.--Experiments with Water .
Glass T. T ' .
t ' .
t. M. 45-0 45-1 45-0 45.4 17'9 17'9 17'9 18-1 o 17-60 17-64 17-64 17-76 14-22 14-25 14-25 14-34 grms. 26'945 26-96 26-965 26-955 1 .
Temperature of the Air 15 ?
'8-16 ?
'1 .
m. f. y. x. sp. H. grs ms .
grs. grm. 11-835 2-095 1-000 0-651 0-0519 , , , , , , , , 0'0519 , , , , , , , , 0-0530 , 2085* , , 0-0542 Mean ... 0-0528 From these determinations , the average of the means of both series of determinations , 0'0518 and 0*0528 , the number 0*0523 is the specific heat of antimony between 17 ?
and 45 ? .
Bismuth.-Purified by melting with nitre , and cast in small bars .
In the case of this metal also , I have made a series of determinations with coal-tar naphtha in the glass , and one with water .
I.--Experiments with Naphtha A. Glass 3 .
Temperature of the Air 180 ?
9-18 ?
'8 .
T. T ' .
t ' .
t. MM .
m. f. y. w. sp. H. oo grms. grms. grm. grm. 50-8 20-6 20-33 18'33 26-99 20-71 1-70 0-431 0-453 0-0291 50-3 20-7 20-42 18-43 26-955 , , , , , , , , 0-0302 50-1 20-6 20-33 18-37 26-955 , , , , , , 0-0292 50-9 20-7 20-40 18-42 26-955 , , 1-685* , , 0-0284 Mean ... 0-0292 II.-Experiments with Water .
Glass 1 .
Temperature of the Air 16 ?
'7-16 ?
*8 .
T. T ' .
t ' .
t. M. m. f. y. X. sp. 1 .
ooo grms. grms. grm. grm. 45-2 18-7 18-44 15-25 26-97 19-43 1-995 1-000 0-651 0-0309 45-5 18-9 18-57 15-36 26-965 , , , , , , .
00313 45-0 18-9 18-64 15-47 26-975 , , , , , , , , 00324 46-0 18-1 18-82 15-56 26-99 1'985 , , , 0-0327 Mean ... 0-0318 From these determinations we get for the specific heat of bismuth between 30 ?
and 48 ?
the number 0-0305 .
36 .
Carbon.-It is known how different are the numbers obtained for the specific heat of carbon in its different forms .
I have determined the specific heat for comparatively only a few of the modifications of carbon-for gas-carbon , for natural and artificial graphite .
Before the experiment each of these substances was strongly heated for some time in a covered porcelain crucible , and then allowed to cool , and immediately transferred into the glass for its reception , and , after weighing , naphtha poured over it .
Gas-carbon from a Paris gas-works ; very dense , of an iron-grey colour , and left very little ash when calcinedt .
It was used in pieces the size of a pea , and two series of experiments were made .
Experiments with Naphtha A. Glass 1 .
Temperature of the Air 18 ' 9-19'O2 ; T. T ' .
tf t. M. m. f. y ... sp. ILY oooo grns .
grms. gr. grm. 52-9 20-8 20'53 18-13 26'955 3-135 1-825 0-431 0'651 0-184 52-6 20-9 20'63 18-26 26-98 , , , , , 04185 51-7 20-7 20-42 18-06 26-97 , , , , , 0-196 52-4 20-9 20-58 18-23 26-98 , 1P805w * , , 0-186 Mean. .
0-188 II.-Experiments with Naphtha A. Glass 3 .
Temperature of the Air 20 ?
-5-20 ?
'8 .
T. T ' .
t ' t. M. m. y sp. ooo grms. g , r g. rm gn .
52'6 22-6 22-33 20-23 26-985 3-345 1-935 0-431 0-453 0-180 52-2 22-5 22-23 20-14 26985 , , , , 0-183 52-3 22-5 22-20 20-12 26-965 , , , , , 0179 52-5 22-6 22'31 20-22 26-955 , , 191* , , 0-182 Mean. .
0181 These determinations give as the average of the means of both sets of experiments the number 0-185 as the specific heat of gas-carbon between 22 ?
and 52 ? .
Natural graphite from Ceylon .
Left very small quantities of ash when calcinedt .
I.-Experiments with Naphtha A. Glass 3 .
Temperature of the Air 18 ?
09-19 ?
02 .
T. T ' t ' .
t. M. m. f. sp. H. ooo0 grms. grms. grs. grm. 51-4 20-8 20-48 18-13 26'975 4-025 2-085 0-431 0'453 0-179 51-4 20-8 20-51 18-13 26-99 , 0-186 51-8 20-8 20'54 18-15 26'975 , , , 0'181 52'0 208 20-54 18-13 2699 , , 206 , 0183 Mean ... 0'183 II.-Experiments with Naphtha A. Glass 1 .
Temperature of the Air 19 ?
'0-18o ?
7 .
T. T ' .
t ' .
t. M. m. .
y. .
sp. H. oooo grs grms. gms .
grm. 53-9 21-1 20-77 18-22 26-97 3-515 1-935 0-431 0-651 0-174 52-2 21-0 20-73 18-31 26'96 , , , , , , , 0-176 52-1 21-2 20-86 18-52 2694 , , , , , , , , 0158 53-0 21-0 20-73 18-32 26-97 , , , , 0.155 52-8 21-0 20-73 18-33 26-965 , 1.91* , , 0.160 Mean ... 0165 III.-Experiments with Naphtha A. Glass 3 .
Temperature of the Air 19 ?
'9-20 ?
'0 .
T. T ' .
t ' .
t. M. m. f. y. .
sp. H. o oo o gros .
grms. grms. grm. 51-6 21-9 21-55 19-33 26-97 3-90 2-05 0-431 0-453 0-174 51-3 22-0 21-71 19-52 26-955 , , , , 0-174 51-5 22-0 21-70 19-52 26'97 , , , , , 0168 51'5 21-9 21-63 19-42 26'96 , , 204* , , , 0-175 Mean. .
0173 The average of the means of these three series of determinations , 0-183 , 0-165 , and 0'173 , gives 0-174 as the specific heat of Ceylon graphite between 21 ?
and 52 ? .
Iron graphite from Oberhammer , near Sayn , separated upon black ordnance iron .
Thin , very lustrous laminae , freed from iron by treatment with aqua regia as much as possible , yet not completelyt .
After drying the stopper .
t This iron graphite , according to Mr. HwBER 's analyses , in which it was also burned in oxygen in a platinum boat placed in a porcelain tube , gave the following results : I. II .
III .
Carbon ... ... . .
97-01 96'12 96-37 Hydrogen ... ... 012 0'18 Ash ... ... ... 4-88 4-87 3-99 101-89 101-11 100-54 It is probable that both in this graphite and in that of natural occurrence , the hydrogen is not essential , but arises from hygroscopic moisture .
The residual ash contained porous particles consisting of sesquioxide of iron and silica , and also small pellets , covered externally with a layer of magnetic oxide of iron : these dissolved in hydrochloric acid at first quietly , and afterwards under disengagement of hydrogen ; and in the solution small blisters of graphite could be perceived .
It is owing to the oxidation of the iron that the sum of the constituents in all cases exceeds 100 .
Experiments with Naphtha A. Glass 3 .
Temperature of the Air 19 ?
'0-18 ?
'7 .
T. 52-9 51'4 52-0 T ' .
t ' .
t. M. grms. 20-8 20a'53 18-21 26-955 21-1 20-84 1854 26-98 20-9 2064 18-43 26-94 20-9 2060 18-33 26-99 m. grms. 2'51 II.-Experiments with Naphtha A. Glass 1 .
T. T ' .
t ' .
t. m. 0o oo o01 Ogrms .
grms. 5241 21-9 21-57 19-32 26-94 2-48 51-7 22-0 21'66 19-45 26-97 51 5 22-0 21-73 19-54 26-98 51-5 22-0 21-66 19'46 26-945 f ... sp. H. grms. grm. 2'445 0-431 0'453 0186 2-565* , , , 0-156 , , , , , 0'157 2-545t , , , , 0-168 Mean. .
0-167 Temperature of the Air 19 ?
'9-20 ?
'0. .
y. x. sp. H. grns .
grm. 2'205 0-431 0'651 0'164 , , , , , , 0-163 , , , , , , 0'162 2-19t , , , 0167 Mean ... 0164 The average of the means of both these series of experiments , 04167 and 0*164 , gives 0'166 as the specific heat of iron graphite between 22 ?
and 52 ? .
The results previously known in reference to the specific heat of carbon , differ greatly for its different conditions , as also do the results obtained by different inquirers and by different methods for the same condition .
But even leaving out of consideration the numbers obtained by DE LA RIVE and MARCET by the method of cooling , there are still considerable differences between REGNAULT'S results , obtained by the method of mixture , and my own .
REGNAULT found for animal charcoal 0'261 , for wood-charcoal 0-241 , for gas-carbon 0-209 , for natural graphite 0-202 , for iron graphite 0-197 , for diamond 0-1469 ; his experiments gave greater numbers for the same substance than my own .
I think that exactly for a substance like carbon in its less dense modifications , my method promises more accurate results than that of REGNAULT .
Even in mine , the substance , after being strongly heated before the experiment , might absorb gases or aqueous vapour , which would make the specific heat too great .
But in REGNAULT'S method this source of error might also operate , and more especially also the source of error due to the disengagement of heat when porous substances are moistened by water .
These sources of error , which affect the determination of the specific heat of the various modifications of carbon and make it too high , have the more influence the looser and the more porous the substance investigated .
I believe that the only certain determination of the specific heat of carbon is that of diamond , and all other determinations are too high , owing to various circumstances , and in REGNAULT'S experiments with wood and animal charcoal , &amp;c. , owing to the heat disengaged when these substances are moistened by water .
Afe soemr.ahh a enadd.1 fe rigtesopr 112 t After drying the stopper .
Chloride of Chromium , Cr2 C16 .
Violet insoluble chloride of chromium twice boiled out with water , washed and dried at 130 ? .
As a porous mass this substance is but ill suited for an accurate determination of the specific heat .
I pressed it , by means of a glass rod , in a glass tube into small disks , between which the naphtha could circulate .
The object of this is to prevent a stagnation of the liquid absorbed by the solid mass , in consequence of which the water of the calorimeter assumes its maximum more slowly , and hence the specific heat is found too low ( compare ? ? ? ?
18 &amp; 24 ) ; but this object is not quite attained in this way* .
Experiments with Naphtha A. Glass 1 .
Temperature of the Air 11 ?
-4-11 ?
'5 .
T. T ' .
t ' .
t. M. m. f. y. .
sp. H. o o~o O grms. grms. grms. grm. 47-5 13-2 12-86 10-32 26-93 3-165 2-095 0-431 0-651 0-139 47-5 13-0 12-73 10-13 26-97 , , , , , 0151 43-8 12-9 12-63 10-33 26-945 , , , , , , 0143 46-0 13-0 12-65 10-21 26'94 , , 2-085t , , , 0140 Mean ... 0143 I should have liked to determine the specific heat of a solid metallic chloride of the formula RCl3 , and tried with chloride of antimony , but it coloured naphtha yellow when poured upon it , and became itself milky white , forming a heavy layer below the naphtha , and fused completely a little above 40 ? .
53 .
Chloride of Zinc and Chloride of Potassium , ZnK2 C1 .
Crystals dried at 100 ?
to 110 ?
$ .
Experiments with Naphtha A. Glass 1 .
Temperature of the Air 14 ' ?
3-14 ?
'5 .
T. T ' .
t ' .
t. M. m. f. y. x. sp. H. oooo grms. grins .
grms. grin .
48-7 16.2 15-93 13-53 26-915 3-01 2-02 0-431 0-651 0-155 47-1 16-3 16-04 13-77 26-965 , , , , , , , 0-155 46-5 16-4 16-12 13-92 26-955 , , , , , , , 0.150 44-1 16'4 16-14 14-13 26-94 , , 2-00t , , , 0-147 Mean ... 0152 1PtFESSOR KO ?
PP ON THE SPECIFIC HEAT OF SOLID BODIES .
and to leave it undecided when , and in how many cases , other such exceptions might occur .
Although the great distance of the temperatures used in my experiments from the melting-point of sesqiuichioride of carbon made it improbable , it was yet possible that the specific heat of this body varies considerably at the temperatures which I used , and is only constant and normal at still lower temperatures .
In the preceding experiments I had heated sesquichloride of carbon to 49 ?
-52 ` ; it was improbable that this body , at so great a distance from its melting-point ( 160 ?
) , should absorb latent heat in softening in appreciable quantity , yet the circumstance that this substance is brittle in the cold , but distinctly tougher at 50 ?
, led me to determine the specific heat at lower temperatures than in the previous case .
I made the two following series of experiments , a with sesquichloride crystallized from alcoholic , and b from ethereal solution : in both series the crystals dried at 1000 were porcelain white in appearance .
a.-Experiments with Water .
Glass. .
Temperature of the Air 17 ?
" 8 .
T. T ' .
t ' .
t. M. m. f. y. x. sp. E. ol oo grms. grms. grms , grin .
36-8 19'7 19-35 17-42 26'98 2-11 2-085 1-000 0-651 0-1446 37'6 19-8 19'52 17'52 26-94 , , , , 0.138 37'2 19'7 19'44 17'51 26-94 , , , , , 0-1I 37'1 19'8 19-45 17-53 26'98 , , 2-075* , , , , 0'147.--Experiments with Water .
Glass 3 .
Temperature of the Air 17 ?
0'8 T. T ' .
t ' .
t M. n. f. y. x. sp. H. 0OO grms. grms. grms. grm. 37-2 19-8 19-45 17-4226-98 , 3-64 2 -11 1000 0-453 0-161 37-2 19-7 19-43 17-42 26-99 , , , , , 0148 37-3 19'7 19-44 1742 26-965 , , , , , , 0-146 37'3 ' 19"7 1944 17-43 26965 , , 210 , , , , 0-145 Both these series can only be considered ajs giving approximate results .
In both the magnitude T T " is very small , ; not as ; much as 18 ?
; in the series a the quantity of solid was : moreover small , and its thermal action ' butt a small fractioni of the entire amount observed .
The mean of the four experiments of the series b would give the specific heat between 20 ?
and 37 ?
at 015 , and !
the first experiment of the series a agrees well with this .
The specific heat kere found between 20 ?
and 37Q comes very near that , calculated from the composition , and is so much less than that found between 20 ?
and ? ?
50 ?
, that it is probable this substance may towards 509 absorb15 heat in softening , the ' atount of which may ' make ' the numbers for the specific heat too great .
To decide upon this point , T made two additional series of experiments in which , since the vessel containing sesqPichloride of carbon and water could , only be slightly heated ( not to 400 ) , and the difference of temperature TT ' accordingly was small , I used all possible care .
I thus obtained the following results .
a. Crystals obtained from ethereal solution dried at 100 ?
: milky white .
Experiments with Water .
Glass 1 .
Temperature of the Air 16 ?
'1-15 ?
*7 .
Tf .
" t ' .
18-1 17-84 18-2 17-92 18-0 17-72 18-2 17-93 18-2 17-93 18-4 18-12 18-1 17-77 t. 15-64 17Q M. grms. 26-94 o^.nn0 m. f. grms. grm. 3-58 1-845 y. 1'000 15-63 26-985 , 1-835* Temperature of the Air 16 ?
'1 .
14-93 26-995 3-58 1'835 1'000 14-95 26-97 Temperature of the Air 16 ?
02 .
13-86 26-995 3-58 1-82 1'000 13-84 26-975 x. sp. H. grm. 0-651 0-174 , , 0176 , , 0-165 0-651.-1 0'193 0'193 0-651 0-269 , , 0-281 b. Clear crystals obtained from ethereal solution , dried air over them at the ordinary temperature .
Experiments with Water .
Glass 3 .
Temperature T. T ' .
t ' .
t. M. m. f. o 1oo 1 10o grms. grms. grms. 36-9 18'2 17-93 15-62 26-99 4-235 2-155 36-8 18-2 17-92 15-64 26-99 , 37-1 18-3 18-01 15-63 26-975 2-145* by passing a current of dry of the Air 16 ?
'2-15 ?
-7 .
y. .
x. sp. H. grm. 1-000 0'453 0-171 0-184 711 ?
0-193 Temperature of the Air 16-1 ?
'-16 ?
-2 .
t. M. m. f. grms. grms. grms. 14-63 26-99 4-235 2-145 14 70I 9.26-9 1 1000 Temperature of the Air 16 ?
'2 .
14-05 26-955 4-235 2-125 1'000 13-73 26-945 , , , ,1 , , x. sp. H. grm. 0-453 0-195 , , 0-195 0-453 0-272 , , 0-285 In the last series of experiments , on heating to about 50 ?
a change took place in the hitherto clear crystals ; they became dull and resembled porcelain .
By special experiments I found that transparent crystals of sesquichloride of carbon gradually heated in water underwent this change at 500-52 ? .
These determinations leave no doubt that , as is the case with other substancest , for temperatures near their melting-points , so also with sesquichloride of carbon at a temperature of 50 ?
( that is more than 100 ?
from its melting-point ) , the specific heat ( or rather the number which is obtained for this in determinations ) rapidly and considerably increases .
From the last two series of experiments the specific heat of sesquichloride of carbon is Between Between Between 18 ?
and 370 .
18 ?
and 43 ? .
18 ?
and 50 ? .
Mean of experiments : a ... 0-172 0-193 0-275 b ... .
0-183 0'195 0-279 Average ... ... ... . .
0-178 0-194 0-277 The specific heat of sesquichloride of carbon increases much more between 43 ?
and 50 ?
than between 37 ?
and 43 ? .
It may be assumed that for temperatures below 37 ?
the number found , 0-178 , comes very near the true specific heat of this compound , that is , uninfluenced by heat of softening .
78 .
Cane-sugar , C12 H2 0 .
Dried crystalline fragments of clear sugarcandy .
Experiments with Naphtha A. Glass 3 .
Temperature of the Air 20 ?
06 .
T. T ' .
t ' t. M. m. f. y. x. sp. H. oooo grms. grms. grm. grm. 49-9 22'2 21-93 19-75 26-96 3-165 1-625 0-431 0-453 0-306 51-4 22-6 22-26 20-03 26-94 , , , , , , 0-295 51-4 22-6 22-30 20-05 26-965 , 1-62* , , , 0-302 Mean ... 0-301 Fine loaf-sugar was recrystallized from water , the mother-liquor washed off with dilute alcohol , the pure white crystals dried at 100 ? .
They gave the following results .
Experiments with Naphtha B. Glass 1 .
Temperature of the Air 18 ?
'5-18 ?
'7 .
T T ' .
t ' .
t. M. m. f. y. .
sp. H. -0 grsgrms grs. grm. grrm .
51-5 20-9 20-62 18'16 26-945 2-915 1-54 0-419 0-651 0-299 51'6 20-7 20-43 17-95 26-95 , , , , , , 0'297 50-3 20-6 20-33 17-94 26-985 , 1-52* , , , , 0-303 Mean ... 0-300 I also examined amorphous cane-sugar .
Crystals dried at 100 ?
, as used in the preceding experiment , were melted in an oil-bath at 160 ?
-170 ?
, and the fused mass allowed to cool in the closed tube .
The resultant amorphous amber-like viscous mass , exactly resembling colophony , was comminuted ( as rapidly as possible to avoid the absorption of moisture ) , and gave the following results .
Experiments with Naphtha B. Glass 1 .
Temperature of the Air 18 ?
'0-1804 .
T. T ' .
t ' .
t. M. m. f. y. .
sp. .
050 grims .
grims .
grin .
grm. 51-4 20-1 19l82 17-24 26-97 2-475 1-77 0-419 0-651 0-336 50-9 20-0 19-74 17-20 26'99 , , , , , , , , 0334 51-6 20'1 19-78 17-15 26-975 , , , , , , , 0-345 50-9 20-1 19-77 17-20 2696 , , 1-75* , , 0-357 Mean ... 0-342 The pieces of amorphous sugar used for these experiments were clear even when the experiments were concluded .
In the investigation of such a hygroscopic substance it is i ?
mpo ; ssible to avoid with certainty any absorption of water ; yet it seems to me improbaile that the difference between the number 0-342 found for amorphous cane-sugar between 20 ?
and 510 , and 0-301 for crystallized sugar between the same limits , depends on ai absorption of water by the former ; but it is probable that the greater specific heat found for amorphous sugar depends on the fact that at 500 even it contains some heat of softening , According to WiHLER'S observations , bodies in the amorphous condition have other , in general lower , fusingspoints than those in the crystallized statet ; crystallized cane-sugar melts at 160 ?
C* , amorphous between 90 ?
and 1000 ; , at the latter temperature the amorphous sugar may be drawn out in threads , but even at a lower temperature the softening begins .
Mannite , Co Hi , 06 .
Crystallized mannite , dried at 100 ?
, was melted in the oil-bath at 160 ?
-170 ?
, and the radiant crystalline mass was comminuted .
It gave the following results 4 .
Experiments with Naphtha B. Glass 3 .
Temperature of the Air 17 ?
'1-17-8 .
T. T ' .
t. t M. m. f. y. x. sp. .
0ooo grms. grins .
grm. grm. 51'1 19.3 18-92 16-57 26-98 2-56 1-815 0-419 0-453 0-318 61-6 19-4 1912 16'64 293 , , , , , , , , 0336 51-0 19-5 19-19 16-82 26-965 , , v 0-319 51-3 19-6 19-31 16'92 26-93 , 1-805 0-321 Mean ... 0'324 * After drying the stopper , t Ann. der Chem. und Pharm. vol. xli p. 155 .
Crystallized Tartrate of Soda and Potass , , H4 Na K +4 H 20 .
Fragments of transparent air-dried Seignette salt , which remained clear in the experiments made with them .
Experiments with Naphtha B. Glass 1 .
Temperature of the Air 16 ?
i7-16 ?
9 .
T. T ' .
t ' .
t. M. .
f. .
x. sp. H. oooo grms. grms grm , grm. 50-0 19-0 18-72 16-03 26*99 3-385 1-415 0-419 0-651 0-324 50'5 18-8 18-47 15-68 26-93 0-333 50'5 18'9 18-57 15-82 26-95 , , , 0 325 50'4 18-9 18-61 15-84 26-965 , , , 0-333 50 5 18'9 18'57 15-83 26'965 , , 140* 0-325 Mean. .
0'328 Crystallized Acid Malate of Lime , G1 H , Ca 05+q4 HG 6 05+8 H1 .
Small crystals dried over sulphuric acid , which remained clear in the following experiments : T. T ' .
t ' .
t. M. m. f. y. sp. E. o0 o~ 0o grms. grms. grm. grm. 50'8 19'4 19-11 16-55 26-985 2-76 1-89 0-419 0-453 0-346 50'1 19'5 19-20 16'73 26-965 , , , , , 0-337 50-5 19'6 19-34 16-84 26-94 , , , , 0'339 50-4 19-6 19-27 16-82 26-97 , 1-865* , , 0-330 Mean ... 0338 IV.-TABLE OF THE SUBSTANCES WHOSE SPECIFIC HEAT HAS BEEN EXPERIMENTALLY DETERMINED .
81 .
In the following I give a summary of those solid substances of known composition for which there are trustworthy determinations of the specific heat .
I have endeavoured to make this summary complete ; yet I have not thought it necessary to include all known determinations ; for instance , all those referring to the metals most frequently investigated .
But it appeared to me desirable to include completely the determinations of experimenters who have investigated a greater number of substances , in order to see how far the results obtained by different inquirers are comparable ; in inserting the numbers which I found for many substances of which the specific heats had been already determined by others , I had no other intention than that of offering criteria for judging how far these determinations are comparable , and may be used for the considerations which are given in the fifth Division .
The determinations given in the following summary are principally due to DULONG and PETIT ( D. P. ) , NEUMANN ( N. ) , REGNAULT ( R. ) , and myself ( Kp .
) .
There are besides some of PERSON ' ( Pr .
) , of ALLUARD ( A. ) , and the recent investigations of PAPE ( Pp .
) are also included .
By far the largest number of these determinations have been made by the method of mixture .
A few only of the elements investigated by DULONG and PETIT , After drying , the stopper .
and some of the chemical compounds by NEUMANN have been determined by the method of cooling .
Where it is not otherwise stated in reference to the temperature , all determinations refer to temperatures between 0 ?
and 1003 .
Where the determination has been made beyond these limits , or where a more accurate statement of temperature is important , it is noticed .
Where the same substance has been repeatedly investigated by the same observer , the result obtained for the purer preparation , and in general the most certain result , is taken .
In the following the chemical formula is given for each substance , the symbols used both here and subsequently , when not otherwise mentioned , refer to the numbers given in the last column of ? ?
2 as the most recent assumptions for the atomic weights , the corresponding atomic weight , and the atomic heat , viz. the product of the specific heat and the atomic weight .
82 .
Elements and Alloys ... ... ... .
27-4 * *** * .
- ... 75 ... .
- ... . .
197 Amorphous. .
9 Graphitoidal .
Crystalline. . ? ? ? ?
... ?
... . .
210 ... ... . .
80 Between 78 ?
and 20 ?
'WVood charcoal .
Gas carbon ... ... . .
12 Natural graphite. .
, .
) ? .
Iron graphite .
Diamond ... ... 112 ... ... ... .- .
58-8 eI Hammered .
j Heated ... 56 j ... 200 .
Between -78 ? ? ? ?
* * ?
a ?
* ?
PROFESSORi KOPP ON THE SPECIFIC HEAT OF SOLID BODIES .
The first formula is that of one constituent of bees ' wax , cerotic acid ; the second is that of the other , palmitate of melissyle .
In reference to the numbers found for the specific heat of bees ' wax at higher temperatures , compare the last remark in ? ?
77 .
Atomic Specific Atomic weight .
heat .
heat .
1 22 11 ... J4 Crystallized cane-sugar ... 0301 Ip .
102*9 f-s^12 2** * for the specific heat of the same body at the same temperature .
At temperatures at which a body softens , the specific heat does indeed vary considerably with the temperature ( compare for example ? ?
77 ) ; but these numbers , as containing already part of the latent heat of fusion , give no accurate expression for the specific heat , and are altogether useless for recognizing the relations between this property and the atomic weight or composition .
Just as little need the small differences be considered iwhich REGNAULT found for a few metallic substances according as they were hammered or annealed , hard or soft .
For dimorphous varieties of the same substance , even where there are considerable differences in the specific gravity , the specific heats have not been found to be materially different ( compare -Fe S , ? ?
83 ; Ti2 , ? ?
85 ; CaC03 , ? ?
86 ) .
The results obtained with these substances appear to me more trustworthy than those with graphite and the various modifications of boron and silicium , which moreover have given partly the same specific heat for the graphitoidal and adamantine modification of the same element .
What trustworthy observations we now possess decidedly favour the view that the dimorphic varieties of the same substance have essentially the same specific heat .
91 .
The view has been expressed that the same substance might have an essentially different specific heat , in the amorphous and crystalline conditions .
I believe that the differences of specific heat found for these different conditions depend , to by far the greatest extent , upon other circumstances .
The Tables in ? ?
83 to ? ?
89 contain a tolerable number of substances which have been investigated both after being melted , and also crystallized ; there are no such differences in the numbers as to lead to the supposition that the amorphous solidified substance had a different specific heat to what it had in the crystallized state .
No such influence of the condition has been with any certainty shown to affect the validity of DULONG and PETIT'S , or of NEUMANN'S law .
I may here again neglect what the determinations of carbon , boron , or silicium appear to say for or against the assumption of a considerable influence of the amorphous or crystalline condition on the specific heat .
REGNAULT found ( ? ?
85 ) that the specific heat of artificially prepared ( uncrystalline ?
) and crystallized titanic acid did not differ .
According to my investigations ( ? ?
48 ) silicic acid has almost the same specific heat in the crystallized and in the amorphlous condition .
In individual cases , where the specific heat of the same substance for the amorphous and crystallized modification has been found to be materially different* , it may be shown that foreign influences affected the determination for the one condition .
Such influences are especially : 1 .
That one modification absorbed heat of softening at the temperature of the experiment ; that is doubtless the reason why the specific heat of yellow phosphorus was found to be considerably greater at higher temperatures than that of red phosphorus , but not at low ones ( compare ? ?
82 ) , that the specific heat of amorphous cane-sugar was found to be decidedly greater than that of crystallized ( ? ?
78 ) , and , according to REGNAULT'S opinion , also that the specific heat of amorphous selenium between 80 ?
and 18 ?
was found much greater ( =0-103 ) than that of the crystalline , while for lower temperatures there was no difference in the specific heats of the two substances ( ? ?
82 ) .
2 .
That in heating one modification its transition into the other is induced , and the heat liberated in this transition makes the numbers for the specific heat incorrect ; in ? ?
33 I have discussed the probability that this circumstance , in REGNAULT'S first experiments with sulphur , gave the specific heat much too high , and it is possible that it was also perceptible in the above-mentioned experiments with amorphous selenium .
3 .
That in immersing heated porous bodies in the water of the calorimeter heat becomes free ( compare ? ?
19 ) ; I consider this as the reason why REGNTAULT found the specific heat of the more porous forms of carbon so much greater than that of the more compact ( compare ? ?
36 ) ; and REGXAIULT himself sees in this the reason why he found the specific heat of the feebly ignited and porous oxides of nickel and of iron greater than that of the same oxides after stronger heating ( compare ? ?
85 ) .
From the importance of this subject for the considerations to be afterwards adduced , I have here had to discuss more fully what differences are real and what are only apparent in the numbers found for the specific heat of one and the same substance .
Even if the apparent differences are often considerable , their importance diminishes , if allowance is made for the foreign influence which may have prevailed .
In many cases , on the other hand , a body in totally different modifications has almost exactly the same specific heat if these foreign influences are excluded .
It may , then , be said that , from our present knowledge , one and the same body may exhibit small differences with certain physical circumstances ( temperature , different degree of density ) , but never so great that they may be taken as an explanation why a body decidedly and undoubtedly forms an exception to a regularity which might have perhaps been expected for it-provided that the determination of the specific heat , according to which the body in question forms an exception , is trustworthy , and kept free from foreign influences .
92 .
Among the regularities in the atomic heat of solid bodies , that found by DULONG and PETIT for the elements stands foremost .
A glance at the atomic heats of the socalled elements collated in 5 82 , shows that for by far the greater number the atomic heats are in fact approximately equal .
But the differences in the atomic heats , even of those elements which are usually regarded as coming under DULONG and PETIT'S law , are often very considerable , even when the comparison is limited to those which are most easily obtained in a pure state , and even if numbers are taken for the specific heats which give the most closely agreeing atomic heats .
REGNAULT * sought an explanation of the differences of the atomic heats of the elements in the circumstance that the latter could not be investigated in comparable conditions of temperature and density ; further , that the numbers for the specific heat , as determined for solid bodies , contain , besides the true specific heat ( for constant volume ) , also the heat : of expansion .
As specific heat we can indeed only take the sum of the heats necessary for heating and , for expansion .
But it is not yet proved that the products of the first quantity ( the specific heat for constant volume ) and the atomic weights would agree better than the atomic heats now do ; it is only a supposition , and even the very contrary may be possible with individual substances .
Temperature has an influence on the specific heat of solid bodies , and to a different extent with different bodies .
Even in this respect , also , all means are wanting by which the different temperatures at which bodies are really comparable can be known , and a comparison made of their atomic heats .
The utmost possible is to determine the specific heat at such a distance from the melting-point that latent heat of softening can have no influence .
It is impossible to say with certainty whether the atomic heats of bodies compared at other temperatures than those which are nearly identical ( ranging about 90 ?
on each side of 10 ?
) will show a closer agreement .
It is not probable .
Changes in the specific heat of solid bodies , so long as they are unaffected by heat of softening , are small in comparison with the differences which the atomic heats of individual elements show .
And it is well worth consideration that individual elements ( phosphorus and sulphur , e.g. ) at temperatures relatively near their melting-points , have not materially greater specific heats than other elements ( iron and platinum , for example ) at temperatures relatively distant from their melting-points , but , on the contrary , considerably smaller. .
As regards the influence of density on the specific heat , it is undoubtedly certain that the latter may somewhat vary with the former ; but it is equally so that , in all cases in which substances of undoubted purity were examined and the sources of error mentioned ( ? ?
91 ) excluded , this variation is too inconsiderable to give an adequate explanation of the differences of the atomic heats found for the various solid elements .
I need not here revert to the considerations developed in ? ? ? ?
90 and 91 , as to how far a difference in the physical condition of a solid substance exercises an essential influence on its specific heat ; for whatever view may be held in reference to this influence , and generally in reference to the circumstances which alter the specific heat of a substance , and therewith the atomic heat , this is certain , that there are individual elements whose atomic heat is distinctly and decidedly different from that of most other elements .
Such elements are , from ? ?
82 , first of all boron , carbon , and silicium .
The decision of the question whether these elements really form exceptions to T)ULONG and PETIT'S law presupposes , besides a knowledge of their specific heat , a knowledge of their atomic weight also .
There can be no exceptions to DULONG and PETIT'S law , if , regardless of anything which may be in opposition to it , the principle is held to , that the atomic weights of the elements must be so taken as to agree with this law .
As a trial whether this law is universally applicable , the atomic weights ought rather to be taken as established in another manner .
It may be confessed that the determination of the true atomic weights by chemical and physico-chemical investigations and considerations is still uncertain , and many questions are still unanswered the settlement of which may influence that determination .
But there seems now to be no more trustworthy basis 180 for fixing the atomic weights of the elements than that of taking , as the atomic weights of the elements , the relatively smallest quantities which are contained in equal volumes of their gaseous or vaporous compounds , or of which the quantities contained in such volumes are multiples in the smallest numbers ; and no better means appear to exist for determining the atomic weights of those elements the vapour-densities of whose compounds could not be determined , than the assumption that in isomorphous compounds the quantities of the corresponding elements are as the atomic weights of the latter .
On this basis , and using this means , the numbers for the atomic weights have been determined which are contained in the last column of the Table in ? ?
2 , and are used in ? ?
82 et seq. The atomic weights B=10'9 , 0=12 , -Si=28 , cannot be changed for others .
That the atomic weights of tin and silicium are as 118 to 28 , is further proved by the isomorphism of the double fluorides .
But to these atomic weights correspond atomic heats which are far smaller than those found for most other elements .
From the chemical point of view it is inadmissible to take the atomic weights of boron , carbon , and silicium * in such a manner as to make their atomic heats agree with DULONG and PETIT'S law .
In any case these three elements form exceptions to DULONG and PETIT'S law .
The sequel will show that this is the case with many other elements .
93 .
In many compounds the regularity is observed , that by dividing their atomic heat by the number of elementary atoms coutained in one molecule of the compound , a quotient is obtained which comes very near the atomic heat of most of the elementsthat is , 6-4 .
This is found in the alloys enumerated in ? ?
82 , and also in a great number of compounds of definite proportions .
A few of the most important cases may be given here .
For speiscobalt , Co As2 ( compare ? ?
83 ) , this quotient is 192 64 ; for the chlorine compounds , R Cl and -R Cltf , the mean of the atomic heats given in ? ?
84 is 12'8 , and the quotient 2-64 .
Of the chlorine compounds , -R C1 , the mean atomic heat of all the determinations in ? ?
84 is 18 5 , and the quotient 5 6 ' 2 .
It is also very near this value in the double chlorides ; inZn K2 Cl4 it is 4 =16'2 , for R K2 C16 ( the mean of the determinations of Pb K2 C16 and Sn K2 C16 ) it is 54'-=6'1 .
For bromine compounds , RBr ( both here and in the following examples the means are taken of the determinations in ? ?
84 ) , l39 -69 ; for Pb Br2 196=65 ; for iodine compounds , RI and RI , i =6 ' 7 , and for the iodine compounds , R I2 1 ' =6 5 .
But this regularity , though met with in many compounds , is by no means quite universal .
The oxygen compounds of the metals correspond to it in general the less the greater the number of oxygen atoms they contain as compared with that of metal .
The mean atomic heat of the oxides Ro in ? ?
85 is 11-1 , and the quotient 1l= 556 .
The quotient for the oxides R03 and R203 ( even excluding the determinations of alumina and boracic acid ) is only 2752 54 ; for the oxides ROU ( even excluding the determinations for silicic acid and zircon ) only 17 = 46 ; for the oxides RO , , the mean of REGNAULT'S determinations only 88s.s47 , Still smaller is the quotient for compounds which contain boron in addition to oxygen ( e.g. for the compounds R B02 ( compare ? ?
87 ) it is only 1 =8= 42 ; for boracic acid , B , 03 , it is only 166 3 ?
3 ) , and also for compounds which contain silicium in addition to oxygen ( it is ll3 8 for silicic acid , Si O , , compare ? ?
85 ) , or which contain oxygen as well as hydrogen ( for ice , H2 0 , it is only 3=2-9* , compare ? ?
85 ) , or which contain hydrogen and carbon besides oxygen ( e.g. it is only 36=2 ' 6 for succinic acid , C H604 , compare ? ?
89 ) .
It may be said in a few words what are the cases in which this quotient approximates to the atomic heat of most elements , and what the cases in which it is smaller .
It is near 6-4 in those compounds which only contain elements whose atomic heats , corresponding to DULONG and PETIT'S law , are nearly =-64 ; it is smaller in compounds which contain elements not coming under DULONG and PETIT'S law and having a much smaller atomic heat than 6*4 , and which are recognized as exceptions to this law , either directly , if their specific heat has been determined for the solid condition ( compare ? ?
92 ) , or indirectly , if it be determined in the manner to be subsequently described .
94 .
The determinations of specific heat given in ? ? ? ?
83 to 89 contain the proofs hitherto recognized for the law that chemically-similar bodies of analogous atomic constitution have approximately the same atomic heat ; and a considerable number of new examples of the prevalence of this regularity are given by my determinations .
The groups of analogous compounds need not again be collated , as NEUMANN has done for a smaller and REGNAULT for a larger number of groups and for individual elements contained in them .
What I will here discuss is the prevalence , beyond the limits of our previous * Considering the atomic heat of liquid water to be 18 , GARNIER ( Count .
Rendus , vol. xxxv .
p. 278 ) thought that the quotient obtained by dividing the atomic weight by the number of elementary atoms in one atom of the compound , J =6 , came near the atomic heat of the elements .
But it requires no explanation that , in a comparison with the atomic heats of solid elements and solid compounds , that atomic heat must be taken for the compound H10 which is obtained from the specific heat of ice , and not from that of water .
GARNIEER is not alone in his error , which is rather to be ascribed to the circumstance that formerly both solids and liquids were compared , as regards their specific heat , in considerations how this property is influenced by the composition .
HERMANN more especially ( Nouveaux Me'moires de la Societe des Naturalistes de Moscou , vol. iii .
p. 137 ) compared liquid water with solid compounds , as did also SCHRODDE ( POGGENDOFFrr 's knowledge , of the regularity , that compounds of analogous atomic constitution have approximately the same atomic heat .
To this belongs , first , the existence of this regularity in the -case of chemically similar bodies , which exhibit an analogy of atomic constitution , when their formulae are written with the atomic weights admitted in recent times for the elements , but which could not be recognized so long as the equivalents of the elements were taken as a basis , or the formula written , as by REGNAULT , with the use of the so-called thermal atomic weights .
The approximate equality of the atomic heats of analogous nitrates and chlorates , of the alkalies for example , had been already observed .
The same character , the haloid , is ascribed both to carbonates and to silicates , but as these formulae were formerly written , an analogy in the composition of chlorates and nitrates , or carbonates and silicates , could not be assumed .
But salts of these four different classes , as well as arseniates and metaphosphates , have analogous atomic constitutions if we assume the recent atomic weights .
The same salts have then also approximately equal atomic heats .
We get the atomic heat Of chlorate of potass , K C1 0 , ? ?
88 ... ... . .
M* 24-8 , , the nitrates , RN03 , in ?
88 ... ... .
M 23-0 , metaphosphate of soda , Na P03 , ? ?
88 ... ... . .
22*1 , , arseniate of potass , KAs03 , ?
88 ... ... .253 .
, , the carbonates , C,3 , ? ?
86 ... ... ... ... M 20*7 , , the silicates , ?
Si0 , 86 ... ... ... 205 The differences in these approximately concordant atomic heats are partly essential and explainable .
I come to this again ( ? ?
95 ) .
According to the more recent assumptions for the atomic weights , certain perchlorates , permanganates , and sulphates have analogous atomic composition , and these salts have also approximately equal atomic heats ; this has been found to be For perchlorate of potass , K Cl0 , ? ?
88 ... ... . .
26 3 , permanganate of potass , K MnO , ? ?
88 ... ... ... 28'3 , the sulphates , SO4 , named in ? ?
88 ... ... .
M 261 But approximate equality in the atomic heat is not only found in such compounds of analogous chemical composition as have similar chemical character , but also in such as have totally dissimilar chemical character .
The chemical character of protosesquioxide of iron ( magnetic iron ore ) is quite different from that of neutral chromate of potass .
Sesquioxide of iron , or arsenious acid , have a chemical character totally different from nitrates or arseniates , or bodies of similar constitution But for the first-named compounds and for the last-named compounds , as respectively compared with each other , there is analegy in chemical composition and approximate equality of atomic heat .
The atomic heat has been found to be For magnetic iron ore , Fe3 O0 , ? ?
85 , , chromate of potass , K2 Gr4 , ? ?
87 , sesquioxide of iron , Fe2 0 , ? ?
85 , , arsenious acid , As2 O , ?
85 M 37-7 M 36-4 M 26-8 .
25-3 , , the nitrates , RNO3 , named in ? ?
88 ... ... 23'0 , , arseniate of potass , KAs03 , ? ?
88 ... ... ... .
25 ' 3 But there is even in a more extended sense approximate equality of atomic heat in bodies of analogous atomic composition .
If the formulae of the oxides , H 02 ( oxide of tin for instance ) are doubled , they become 2 04 , and are then analogous to those of the sulphates , RS 04 , or of tungstate of lime or of perchlorate of potass and other salts .
To the formulae thus made analogous equal atomic heats correspond .
The following:atomic heats have been found : Oxide of tin , Sn2 04 , compare ? ?
85 ... ... .
Titanic acid , Ti2 0 , , , The sulphates , R SO4 , in ? ?
87 ... ... . .
M 27-6. .
M 27-3. .
M 26-1 Tungstate of lime , Ca W 04 , compare ? ?
87 ... ... ... 279 Perchlorate of potass , K Cl 04 , compare ? ?
88 ... ... . .
263 Permanganate of potass , K Mn O4 , compare ? ?
88 ... ... .
28-3 If the formulae of the oxides , R-O2 , are trebled they become R3 O6 , analogous to those of the nitrates H Ns 06 ( nitrate of baryta , e. g. ) , and similar salts .
Here also approximately equal atomic heats correspond to the formulae thus made analogous .
The atomic heats are as follows : Oxide of tin , Sn3 06 , compare ? ?
85 ... ... ... .
M 41'4 Titanic acid , Ti3 0 , , , .
The nitrates , X N2 06 , in ? ?
88 ... Metaphosphate of lime , Ca P2 06 , compare ? ?
88. .
M 41-0. .
M 38-1 ... . .
39-4 How little the atomic heat of compounds depends on their chemical character may be proved from a greater series of examples than those adduced in the preceding .
It is , however , unnecessary to dwell upon this .
The comparisons and considerations contained in the sequel complete what has here been developed as a proof of the principle that the atomic heat of bodies is independent of their chemical character .
95 .
The foregoing comparisons give examples of cases in which bodies of analogous atomic structure , with a totally different chemical character , have approximately the same atomic heat ; they show that with reference to the atomic heat , monoequivalent and polyequivalent elementary atoms have the same influence , which , indeed , followed already from REGNAULT'S comparisons ; that the atomic heat of a substance for its polyfold atomic formula may be compared with that of another substance for a simple atomic formula .
184 The preceding contains a generalization of NEUMANN'S law ; but as certainly as this law is recognized in the preceding in a more general manner than was formerly assumed , as little is it universally applicable .
REGNAULT'S investigations have shown that NEUMANN'S law is not rigidly valid .
Even for those compounds which contain the same element as electronegative constituent , and have similar atomic constitution , he found the atomic heats as much as no to -1 different from each other* .
The reason of this he seeks in the same circumstances , which in his view prevent a closer agreement in the atomic weights of the elements ( compare ? ?
92 ) .
Differences of this kind , and even still more considerable , occur in the atomic heats of compounds for which greater agreement in these numbers might be expected-of such compounds , rthat is , as contain elements of the same , or almost the same atomic heat combined with the same other element in the same atomic proportion. .
To this belongs the fact that the atomic heat has been found so different ( ? ?
85 ) for the isomorphous compounds , magnetic iron ore ( 37-7 ) , chrome iron ore ( 31*2 ) , and spinelle ( 27-7 ) , and for alumina ( 21 3 ) and for sesquioxide of iron ( 26 8 ) .
In the atomic heats of such analogous compounds there are differences for which , or rather for the magnitude of which , as furnished by our present observations , I know at present no adequate explanation .
But there is another kind of difference in the atomic heats of analogous compounds , which exhibits a regularity , and for which an explanation can be given .
Certain elements impress on all their compounds the common characteristic , that their atomic heat is much smaller than that of most analogous compounds .
The atomic heat of boracic acid , B 03 , is only 16*6 , while that of most other compounds , R2 O3 and R2 03 , is between 25 and 28 ( ? ?
85 ) .
The atomic heat of the borates , RB 02 , is ( ? ?
87 ) only 16'8 , while that of R2 -2 , as the mean of the determinations in ? ?
85 , is 22-2 .
The atomic heat of Pb B2 04 is ( ? ?
87 ) only 26-5 , while that of Fe3 04 ( ? ?
85 ) in the mean is 37 7 .
Similar results have been obtained for compounds of certain other elements , of carbon and of silicium for instance , that is , of those elements which in the free state have a smaller atomic heat than that of most other elements .
This observation leads to the question whether the elements enter into compounds with the atomic heats which they have in the free state , and in connexion with this , how far is it permissible to make an indirect determination of the atomic heat of the elements ( in their solid state ) from the atomic heats of their ( solid ) compounds .
96 .
The assumption that elements enter into compounds with the atomic heats they have in the free state would be inadmissible , if not only the atomic structure as expressed by the empirical formula , but also the grouping of the elements to proximate constituents , as is endeavoured to be expressed by the rational formula , influenced the atomic heat of the compounds .
That the latter is not the case is very probable from the comparisons made in ? ?
94 , where approximately equal atomic heats were obtained for compounds of analogous empirical formulae , even with the greatest dissi* Ann. de Chim .
et de Phys. [ 3 ] vol. i. p. 196 .
2c2 185 milarity of chemical character .
That that , which may be supposed and expressed by the so-called rational formula in reference to the internal constitution of compounds , does not affect the atomic heat , becomes more probable from the fact that chemically similar , and even isomorphous compounds , one of which contains an atomic group in the place of an individual atom in the other , exhibit dissimilar atomic heats .
This is seen , for instance , in comparing analogous chlorine and cyanogen compounds ( Cy=CN ) ; the latter have far greater atomic heats .
Thus the atomic heat Of chloride of mercury , H-g C12 , 84 , is ... ... ... .
18-0 cyanide of mercury , Hg Cy2 , ? ?
89 ... ... ... . .
25*2 , chloride of zinc and potassium , Zn K2 C14 , ? ?
84 ... ... .
43'4 , , cyanide of zinc and potassium , Zn K2 Cy , ? ?
89 ... ... .
59*6 In like manner ammonium compounds ( Am=N H4 ) have atomic heats considerably greater than the corresponding potassium compounds .
This is seen from the following Table : Chloride of potassium , K Cl , ? ?
84. .
, , ammonium , Am C1 , ? ?
84 .
Nitrate of potass , KN 03 , ? ?
88. .
, , ammonia , Am N 03 , ? ?
88 Sulphate of potass , K2 S04 , ? ?
87 .
, , ammonia , Am2 804 , ? ?
87 ... . .
M 12'9 ... ... 20-0 ... M 23-5 ... ... .
364 ... ... ... M 33-6 ... ... .
462 97 .
That undecomposable atoms and atomic groups are contained in compounds with the atomic heats they have in the free state is further probable from the fact that the ' sum of the atomic heats of such atoms , or atomic groups , as when united form a certain compound , is equal or approximately equal to the atomic heat of this compound .
For many compounds whose elements obey DULONG and PETIT'S law , what has been stated in ? ?
93 contains the proof that the atomic heat of these compounds is equal to the sum of the atomic heats of the elementary atoms contained in one atom of the compounds .
That this is also observed when atomic groups are supposed to be united , forming more complicated compounds , will be seen by bringing forward a few examples .
The atomic heat has been found For the oxides , R O , enumerated in ? ?
85 ... . .
, , sesquioxide of iron , Fe2 3 , ? ?
85. .
Sum for Fe2 R04 , , magnetic iron ore , Fe3 0 , ? ?
85 ... ... , the oxides , -R O , in ? ?
85 ... , , the acids , R 03 , in ? ?
85 , according to REGNAULT Sum for , , chromate of lead , Pb Gr 04 , ? ?
87 ... . .
M 11.1 M 26-8. .
337.9 .
M 37.7. .
M 11-1 ... I. M 18-8 B O4 ... 29'9 ... ... 29'0 186 The Tables in ? ?
84 to 89 contain data for several such comparisons , which lead to the same result as the preceding-that the atomic heat of water contained in solid compounds may , by subtracting the atomic heat of the anhydrous solid from that of the hydrated solid compound , be obtained in sufficient approximation to the atomic heat deduced from the direct determination of the specific heat of ice .
The deviations from each other and from the atomic heat of ice as directly determined , which these indirect determinations exhibit , are not to be wondered at when it is considered that all uncertainties in the atomic heats , from whose difference the atomic heat of solid water is deduced , are concentrated upon this difference .
98 .
The view already expressed and defended ( compare especially ? ?
12 and 13 ) , that atoms and atomic groups are contained in solid compounds with the same atomic heat which they have in the free state , is opposed to the view which has also been frequently expressed and defended-that the atomic heat of an element may in certain compounds differ from what it is in the free state , and may be different in different compounds .
This view , and the reasons which may possibly be urged in its favour , must here be discussed .
The first statement of this view ( compare ? ?
6 ) simply goes to assert that the atomic heats of compounds may be calculated in accordance with the values resulting from the determinations of the specific heat , assuming that one constituent of the compound has the same atomic heat as in the free state , the other an altered one .
What alteration is to be assumed depends merely on what assumption adequately satisfies the observed specific heat of the compound .
The accuracy of the assumption is susceptible of no further control ; the assumption itself cannot be regarded as an explanation of the observed atomic heat of the compound .
And nothing is altered in this by assuming ( compare ? ?
6 and 11 ) that the changes in the atomic heat of a substance on entering into chemical compounds take place in more or less simple ratios .
A greater degree of probability must be granted to the view ( compare ? ?
10 ) that the atomic heats of the constituents of compounds , and the differences in the atomic heats of these bodies , according as they are combined or in the free state , depend upon the state of condensation in which these bodies are contained .
If , for instance , from a consideration of the specific gravities or specific volumes ( the quotient of the specific weights into the atomic weights ) of compounds and of their constitutents , a conclusion could be drawn with some degree of certainty as to the state of condensation in which the latter are present in the former , and if definite rules could be given for the variations of the atomic heats with the state of condensation , the result of such an investigation , if it agreed with the observed results for the atomic heats of compounds , might be called an explanation of these observations .
But what is here presupposed is partially not attained and partially not attempted .
And , moreover , as far as can be judged from individual cases , the same element , when contained in different states of condensation , appears to have the same atomic heat .
It has been attempted to deduce the state of condensation , or the specific volume of oxygen in its compounds with heavy metals , 188 by subtracting from the specific volume of the oxide that of the metal in it , and considering the remainder as the volume of oxygen .
It would follow from this that the specific volume of oxygen in suboxide of copper is much greater ( about four times as great ) than in oxide of tin .
But if the atomic heat of oxygen be deduced by subtracting from the atomic heat of the oxide that of the metal in it , it is found that the atomic heat of oxygen in suboxide of copper and in oxide of tin gives almost exactly the same number .
Hence it does not seem that the state of condensation in which a constituent may be contained in a compound has any material influence on the atomic heat of this constituent .
99 .
From all that has been said in the foregoing paragraphs the following must be adhered to .
( 1 ) Each element in the solid state , and at a sufficient distance from its melting-point , has one specific or atomic heat , which may , indeed , somewhat vary with physical conditions , different temperature , or density for instance , but not so considerably as to be regarded in considering in what relations the specific heat stands to the atomic weight or composition ; and ( 2 ) that each element has essentially the same specific or atomic heat in compounds as it has in the free state .
On the basis of these two fundamental laws it may now be investigated what atomic heats individual elements have in the solid free state and in compounds .
Indirect deductions of the atomic heats of such elements as could not be investigated in the solid free state are from these propositions admissible : that from the atomic heat of a compound containing such an element the atomic heat of everything else in the compound is subtracted , and the remainder considered as the expression for the atomic heat of that element .
Such indirect determinations of the atomic heat of elements may be uncertain , partly because the atomic heat of the compounds is frequently not known with certainty , as is seen from the circumstance that analogous compounds , for which there is every reason to expect the same atomic heat , are found by experiment to have atomic heats not at all agreeing ; but more especially because the entire relative uncertainty in the atomic heats for a compound , and for that which is to be subtracted from its composition , is concentrated upon a small number , the residue remaining in the deduction .
But when such deductions are made , not merely for individual cases , but for different compounds , and for entire series of corresponding compounds , they may be considered sufficiently trustworthy to make the speculations based upon them worthy of attention .
Of course in indirectly deducing the atomic heat of an element , its simpler compounds , and those containing it in greatest quantity ( measured by the number of atoms ) , promise the most trustworthy results .
100 .
For Silver , Aluminium , Arsenic , Gold , Bismuth , Bromine , Cadmium , Cobalt , Copper , Iron , Jlercury , Iodine , Iridium , Potassium , Lithium , iMagnesium , Manganese , Molybdenum , Sodium , Nickel , Osmium , Lead , Palladium , Platinum , Bhodium , Antimony , Selenium , Tin , Tellurium , Thallium , Tungsten , and Zinc , it may be assumed , from the determinations of their specific heat in the solid state ( ? ?
82 ) , that their atomic heats , in 189 accordance with DULONG and PETIT'S law , are approximately equal , the average being 6-4 .
I do not think that all these elements have really the same atomic heat , but think that some of them will subsequently be considered as exceptions to the above-mentioned law , as it will in the sequel be proved that several elements have an atomic heat differing from 6*4 .
But for none of the previously mentioned elements are the present data , and the presumed deviation of the atomic heat from that of other elements , sufficient to justify their being separated from them .
To the elements just mentioned chlorine must be associated from the close agreement of the corresponding chlorine , bromine , and iodine compounds ( ? ?
84 ) , and of the compounds K Cl 03 , 24-8 , and K As 03 , 25'3 ( ? ?
88 ) .
To the atomic heats of these latter compounds those of individual salts KN 03 approximate closely ; the latter gave ( ? ?
88 ) 21-8-24-4 , mean 23'0 , which on the whole agrees sufficiently closely with those found for the metallic oxides , R2 03 ( ? ?
85 ) .
I count nitrogen also among the elements whose atomic heat may be assumed at 6*4 , like that of most other elements ; without , however , considering the determination of the atomic heat of this element as very trustworthy .
To deduce the atomic heat of this element with certainty , compounds are wanting which contain , besides nitrogen , elements whose atomic heat has been directly determined .
The fact that the atomic heat of the nitrates , RB N2 0 , was found ( ? ?
88 ) in the mean to be 38X1 , a third of which , 127 , is somewhat less than the average atomic heat found for the oxides of heavy metals of the formula R 02 , might be a reason for assigning to nitrogen a smaller atomic heat ; while , on the other hand , the atomic heats of other nitrogen compounds , in which it is true other elements enter whose atomic heat is only indirectly determined , do not favour this view .
In the class of elements with the atomic heat about 6-4 , barium , calcium , and strontium may be placed from the agreement in the atomic heats of their compounds with the atomic heats of corresponding compounds of such elements as have been found by the direct determination of their specific heat in the free solid state to belong to that class ( compare the atomic heats of the compounds C1l in ? ?
84 , C 03 in ? ?
86 , RSO4 in ? ?
87 , and RN206 in ? ?
88 ) ; further , rubidium ( compare the atomic heats of the compounds R C1 in ? ?
84 , and R2 C 03 in ? ?
86 ) ; then also chromium ( from the agreement in the atomic heats of Gr2 03 and Fe2 03 , ? ?
84 ) , and titanium ( from the agreement in the atomic heats of Ti 02 and Sr 02 , ? ?
84 ) .
To place zirconium in the same class has no other justification than that on this assumption the atomic heat of zircon may be calculated in accordance with that deduced from the observed specific heat of this mineral .
101 .
According to direct determinations of the specific heat , sulphur and phosphorus do not belong to this class .
The more trustworthy determinations ( for sulphur the last two , for phosphorus the last three of the numbers in ? ?
82 ) assign to these elements the atomic heat 5*4 .
That sulphur has a smaller atomic heat than the elements discussed in the last paragraphs follows from the atomic heats of sulphur compounds , compared 190 with those of the corresponding compounds of such elements as have an atomic heat =6'4 .
The average atomic heat of compounds RS and SS is 11'9 , according to the determinations in ? ?
83 , while those of chlorine compounds R C1 and R C1 ( ? ?
84 ) =12'8 , that of the corresponding bromine compounds =13'9 , and of the corresponding iodine compounds =13'4 .
In comparing more com'plicated sulphur compounds , sulphates , for instance , with other compounds of analogous composition , the same is met with ; although such complicated compounds are of little value in giving data for deciding on such small differences .
The specific heat of the simpler phosphorus compounds has not been investigated ; for more complicated compounds , although they point to a smaller atomic heat for P than 6'4 , the above remark also applies .
The determinations of the specific heat of silicium give for this element also a smaller atomic heat than 64 ( compare ? ?
82 ) , and the same conclusion results from a comparison of the atomic heats of Si 02 , and the oxides , R 02 , of the silicates R Si 03 , and the oxides R2 03 .
The atomic heat to be assigned to silicium cannot as yet be settled with any degree of certainty .
Direct determinations , varying considerably from each other , give a specific heat mostly greater than 4 ; while the numbers obtained indirectly , and themselves also not closely agreeing , are partly considerably smaller .
If in the sequel I put the atomic heat of silicium at 3*8 , corresponding to the lowest number found for the specific heat of this element , I do so for want of other and more certain data .
I consider this number as quite uncertain .
The atomic heat of boron , from the direct determinations of the specific heat , is considerably smaller than 6*4 ; and the atomic heats of boron compounds confirm this , as was discussed in ? ? ? ?
93 and 95 .
By comparing the atomic heats of such boron and sulphur compounds as contain along with boron and sulphur the same elements in the same proportions , the atomic heat of boron is found to be half that of sulphur .
The atomic heat of KB 03=16*8 is exactly half that found for K2 S 04= =33 -6 ; the atomic heat of Pb B2 04=26'5 is almost exactly equal to that for P-b SO4=25'7 .
Taking the atomic heat of S , in accordance with the above discussion , at 5'4 , that of B would be 2'7 ; the numbers obtained directly for the atomic heat of boron ( ? ?
82 ) from the experiments on the specific heat of this element agree with sufficient accuracy .
In the sequel I take the atomic heat of B at 2'7 .
A smaller number is obtained in other comparisons ; for instance , of the atomic heats of B2 03 and of the oxides X2 03 , or of the salts RB 02 and the oxides R2 02 ; but in such indirect determinations of the atomic heat , where such small numbers are to be determined , as is here the case with the atomic heat of boron , the results are very uncertain , owing to the fact that the entire uncertainty in the atomic heats of the compounds , and in the assumption that the elements corresponding to boron in compounds of analogous composition have really the atomic heat =6'4 , is thrown on the final result .
Lastly , carbon also , from the direct determinations of its specific heat ( ? ?
82 ) , has a much smaller atomic heat than 6'4 .
The same result follows from a comparison of the atomic heats of carbon compounds : the atomic heat of the carbonates , -R2 C 03=28-4 as the mean of the determinations in ? ?
86 , is much smaller than that of R303(=3BO ) , which is the mean of the numbers in ? ?
85 =33'3 ; the atomic heat of the carbonates RCO3 =20'7 , as the mean of the determinations in ? ?
86 , is much smaller than 27*1 , the number found for As2g , Bi2 0 , Cr20O , Fe2 0 , and Sb2 03 as the atomic heat of oxides -2 03 ' I put the atomic heat of carbon at 1'8 for f , as deduced from the determination of the specific heat of its purest variety , diamond .
102 .
In the preceding paragraphs I have discussed the elements which , from the determinations of their specific heat in the solid free state , have a smaller atomic heat than about 6 4 .
There remain to be discussed a few elements whose atomic heats are also less than those of most other elements , but can only be deduced from those of their compounds .
To this category belongs hydrogen* , even if the indirect determination of its atomic heat in the solid state is liable to the uncertainty just discussed .
The atomic heat of water , 1H O , is ( ? ?
85 ) =8-6 , and smaller by 7 than that of suboxide of copper , Cu2 O , which was found in the mean to be 15'6 ; the atomic heat of hydrogen would thus be j-=3*5 less thain that of the elements to which copper belongs , as regards its atomicheat ; hence the former would be 6435= 2 ' 9 .
The atomic heat of chloride of ammonium , N HI Cl , has been found to be 20 0 ( ? ?
84 ) ; the subtraction of the atomic heats for N+-GCl=z64+6'4-=12'8 , leaves 7'2 as the atomic heat of 4H , and therefore 1'7 for that of H. The atomic heat of nitrate of ammonia , N 114 03 , is 36-4 ( ? ?
88 ) ; subtracting therefrom as the atomic heat of N2+OS , the number 27'1 , which has previously bteen frequently mentioned as the atomic heat of oxides -BR 03 we have 9-3 as the ato-mic heat of 4H , that is 2'3 for that of H. I put in the sequel the atomic heat of hydroyen at 2-3 .
That oxygen has a smaller atomic heat than 6'4 , follows from the fact that the oxygen compounds of the metals have a considerably smaller atomic heat than the corresponding chlorides , iodides , or bromides .
For instance , the atomic heat of the oxides -I 0 is as the mean of the determinations in ? ?
85 -=11 1 , while that of the chlorides R C1 and -I ) C1 ( ? ?
84 ) , is 12'8 , that of the corresponding bromides 13'9 , and of the corresponding iodides 13-4 .
That of the oxides , R 02 , as the mean of the determinations in ? ?
85 , of Mnn2 , -Sn 0-2 , and Ti O2 is 13'7 , while that of the chlorides RC12 ( ? ?
85 ) is 18'5 , and of the iodides i:I 2=19'4 .
Taking the atomic heat of the other elements , which are contained in the following compounds , at 6'4 , the atomic heat of oxygen , as deduced from the atomic heat of the oxides -R 0 ( 11-1 in the mean ) , is =4-7 ; as deduced from the oxides R2 03 ( 27-1 as the mean of the oxides of this formula previously frequently mentioned ) , it is =4-8 ; from the above oxides , B0(13 7 in the mean ) , it is =3'7 ; it is found ( compare ? ?
88 ) from K As 03 ( 25-3 ) to be 4-1 ; from Pb3 As2 0 ( 65-4 ) to be 4'2 ; from K Cl Og ( 24-8 ) to be 4'0 ; from K Cl 04 ( 26-3 ) to be 3-4 ; from K Mn 04 ( 28-3 ) to be 3-9 .
In the sequel I take the round number 4 for the atomic heat of 0 .
L. GMELIN ( Handbuch der Chemie , 4 Aufl .
vol. i. pp. 216 and 222 ) ascribed to hydrogen the saime capacity for heat as that of an equivalent quantity of lead or mercury ( H==1 , Cu=31'7 , Hg= 100 ) ; ScnRO , ER ( PoaGreo .
Ann. vol. lii .
p. 279 ) and CANNIZZARO ( I1 Nuovo Cimento , vol. vii .
p. 342 ) ascribed to hydrogen the same atomic heat as that of most other elements ( _I=l , C1=35'5 , &lt; u=63'4 , Hg=200 ) .
192 Fluorine appears , lastly , to have a considerably smaller atomic heat than 6'4 .
The atomic heat of fluoride of calcium , -a Fl12 has been found to be ( ? ?
84 ) only 16 4 , considerably smaller than the corresponding chlorides , bromides , and iodides .
I put the atomic heat of fluorine at 16 4-64=5 .
103 .
Taking , in accordance with what has just been said , the atomic heat which an element has in a solid compound , At 6-4 for Ag , Al , As , Au , Ba , Bi , Br , Ca , Cd , C1 , Co , Cr , Cu , Fe , -Hg , I , Fr , K , Li , Mg , Mn , Mo , N , Na , Ni , Os , Pb , Pd , -t , Rb , RJh , Sb , Se , -Sn , Sr , Te , Ti , Tl , W , Zn , and Zr , At 5-4 for S and P , at 5 for Fl , 4 for 0 , 3-8 for Si , 2-7 for B , 2-3 for H , and 1-8 for ?
; and assuming that the atomic heat of a solid is given by the sum of the atomic heats of the elements in it , we obtain the atomic heats ; and dividing them by the atomic weights , we obtain the specific heats , in sufficiently close agreement with the specific heats as obtained by direct determinations of this property .
In the following Table I give for all compounds for which the specific heat has been determined in a trustworthy manner , the specific heat calculated on these assumptions , compared with the numbers found experimentally .
I give this calculation and this comparison in the same order which was followed in the synopsis ? ?
82 to 89 , and I refer to the latter as regards special remarks on the determinations .
To distinguish the observers , N. again stands for NEUMANN , R. REGNAULT , Kp .
KoPP , Pr .
PERSON , A. ALLUARD , and Pp .
PAPE .
Bi Sn .
BiS n2 .
Bi Sn2 Sb Bi Sn2 , Sb Zn2 Pb Sb. .
PbSn. .
Pb Sn2 .
Co As .
Ag2 S. Co AsS .
CuS. .
Fe As S AsS .
Co S. .
Cuo-Fe-S. .
Fe -S .
HgS .
NiS. .
Alloys .
( Compare ? ?
82 .
) A Atomic Specific Specific Atomic heat .
heat .
heat .
weight .
Calculated .
Calculated .
Observed. .
328 12-8 0-0390 0-0400 R. .
446 19-2 0-0430 0-0450 R. .
568 25-6 0-0451 0-0462 R. .
698-4 38-4 0'0550 00566 R. .
329 12-8 0,0389 0-0388 R. .
325 12-8 0-0394 0-0407 R. .
443 19-2 0'0433 0-0451 R. 104 .
Arsenides and Sulphides , ( Compare ? ?
83 .
) .
z 2. .
I ... .I1 ?
..12 !
08-8 19-2 0-0919.00920 '48 18-2 00734 0-0746 66 18-2 0-110 0'107.58-8 182 0-115 0121.63 18-2 0-112 0'101.07 11'8 0'110 0.111 90'8 118 0'130 0-125 91-7 11-8 0-129 0-129 88 11-8 0-134 0-136 132 11-8 0-0509 0-052 90-8 118 0-130 0-128 22 N. R. N. R. 0-120 N. N. R. N. 0-131 R. N. 0-051 : R. Kp .
Kp .
2 R. 0'0517 Kp .
193 Atomic Atomic Specific Specific Ateght heat .
heat .
heat .
weight .
Calculated .
Calculated .
Observed .
QuadroxalateofpotC2H3K08+2H. .
254 ' 1 69'7 0274 0-283 Kp .
ass Bitartrate of potass CH KO6 ... 188-1 49*1 0-261 0-257 Kp .
Seignette salt. .
C4H4NaKO6+4H0 282-1 87-6 0-311 0-328 Kp .
Bimalateofpotass .
C H10 Ca 10-+8 12 0 .
450 152-6 0-339 0-338 Kp .
111 .
The preceding synopsis shows , for the great majority of substances contained in it , an adequate agreement between the observed specific heats and those calculated on such simple assumptions .
In estimating the differences , the extent must be remembered to which various observers differ for the same substance .
It must be considered that the present better determinations of the specific heat , even those made by the same experimenter , for substances where it may be expected that NEUMANN'S law applies , do not exactly agree with it , not more nearly than within 1or ~ of the value ; and that for those elements which are considered here as obeying DULONG and PETIT'S law , even greater deviations occur between the numbers found experimentally and those to be expected on the assumption of the universal validity of this law .
( These deviations , i. e. the differences between the atomic heats found for these elements , are seen from ? ?
82 .
) The extent to which the experimentally determined specific heats deviate from such a law , NEUMANN'S for instance , in bodies for which calculation takes it as applying , gives of course the means of judging what differences may occur between the observed and calculated numbers without invalidating the admissibility of the calculation attempted .
And it is as much a matter of course that , in those bodies in which a marked deviation from NEUMANN'S law has been already mentioned ( compare ? ?
95 ) , a greater difference is found in the present synopsis between calculation and observation .
I consider the agreement between calculation and observation , as shown in the synopsis ? ?
103 to 110 , as in general sufficient for a first attempt of that kind .
But it need scarcely be mentioned that I by no means consider the calculated as more accurate than the observed numbers , or among sevexal numbers consider that the most accurate which is nearest the calculated ; for that , the bases of calculation are much too uncertain .
The list of atomic heats given at the commencement of ? ?
103 is scarcely much more accurate than were the first tables of atomic weights ; but just as the latter have experienced continual improvements , and thus what was at first only an approximate agreement between the calculated and observed composition of bodies has been brought within considerably narrower limits , and apparent exceptions been explained , so , in like manner , will this be the case for ascertaining what atomic heats are to be assigned to the elements , and how the atomic heats of compounds may be deduced therefrom .
This much , how.ever , may even now be said , that while formerly for many solid substances a statement of the specific heat could in no way be controlled , a concealed source of error for the determination of this property was not indicated , and an error which materially altered the number for this property could not be recognized , at present , even if only roughly , such a control is possible .
Compare ? ?
77 .
198 The proof given in the preceding that DULONG and PETIT'S law is not uniuersally valid , justifies certain conclusions , in reference to the nature of the so-called chemical elements , which may here be developed .
What bodies are to be regarded as chemical elements ?
Does the mere fact of indecomposability determine this ?
or may a body be indecomposable in point of fact and yet from reasons of analogy be regarded not as an element but as a compound ?
The history of chemistry furnishes numerous examples of cases in which sometimes one and sometimes another mode of view led to results which at present are regarded as accurate .
The earths were in 1789 indecomposable in point of fact , when LAVOISIER expressed the opinion that they were compounds , oxides of unknown metals .
LAVOISIER'S argumentation was based on the fact that the earths enter as bases into salts , and that it was to be assumed in regard to all salts , that they contained an oxygen acid and an oxygen base .
But the view , founded on the same basis , that common salt contains oxygen , and the subsequent view that what is now called chlorine contained a further quantity of oxygen besides the elements of an oxygen acid , did not find an equally permanent recognition .
On the basis of the actual indecomposability of chlorine , DAVY maintained from about 1810 its elementary character ; and this view has become general , especially since BERZELIUS , after a long struggle against it , adopted it , more I think because he was outvoted than because he was convinced .
Almost all chemists of the present time consider chlorine , and in conformity therewith bromine and iodine , as elementary bodies ; but the persistence is known with which SCHONBEIN attacks this view , and adheres to the opinion that these bodies are oxygen compounds , peroxides of unknown elements .
Is there anything which enables us to decide with more certainty on the elementary nature of chlorine and the analogous bodies than has hitherto been the case ?
No one can maintain that the bodies which chemists regard as elements are absolutely simple substances .
The possibility must be confessed that they may be decomposed into still simpler bodies ; how far a body is to be regarded as an element is so far relative , that it depends on the development of the means of decomposition which practical chemistry has at its disposal , and on the trustworthiness of the conclusions which theoretical chemistry can deduce .
A discussion as to whether chlorine or iodine is an elementary body can only be taken in the sense whether chlorine is as simple a body as oxygen or manganese , or nitrogen ; or whether it is a compound body , as peroxide of manganese or peroxide of hydrogen for example .
If DULONG and PETIT'S law were universally valid , it would not merely indicate for chemical elements a relation between the atomic weight and the specific heat in the solid state , but it could be used as a test for the elementary nature of a body whose atomic weight is known .
That iodine , from a direct determination of specific heat , and ' chlorine by an indirect determination had atomic heats agreeing with DULON ' and law , would be a proof that iodine and chlorine , if compounds at all , are not more so than other so-called elements for which this law is regarded as valid .
According to NEUMANN'S law , compounds of analogous atomic composition have approximately the same atomic heats .
In general , bodies , whose atom consists of a greater number of indecomposable atoms , or is of more complicated composition , have greater atomic heats .
In these compounds , more especially those whose elements all follow DULONG and PETIT'S law , magnitude of atomic heat is exactly a measure of the complexity or of the degree of composition ( compare ? ?
93 ) : .
If DULONG and PETIT'S law were valid , it could be concluded with great positiveness that the so-called elements , if they are compounds of unknown and simpler substances , are compounds of the same order .
It would be a remarkable result that the act of chemical decomposition had everywhere found its limit at such bodies as those which , if compound at all , have with every difference of chemical deportment the same degree of composition .
Imagine the simplest bodies , probably as yet unknown to us , the true chemical elements , forming a horizontal spreading layer , and piled above them , the simpler and then the more complicated compounds ; the universal validity of DULONG and PETIT'S law would include the proof , that all elements at present assumed by chemists lay in the same layer , and that chemistry in recognizing hydrogen , oxygen , sulphur , chlorine , and the different metals as indecomposable bodies , had penetrated to the same depth in that field of inquiry , and had found at the same depth the limit to its penetration .
This result I formerly propounded * when I still believed in the validity of DULONG and PETIT'S law .
But with the proof that this law is not universally true , the conclusion to which this result leads loses its justification .
Starting now from the elements recognized in chemistry , we must rather admit that the magnitude of the atomic heat of a body depends not oly on the number of elementary atoms contained in one atom of it , or on the compleatomic hxieat of the composition , but also on the atomic heat of the elementary atoms entering into its composition ; it appears now possible that a decomposable body may have the same atomic heat as an indecomposable one .
To assume in chlorin the the presence of oxygen , and to consider it as analogous to peroxide of manganese , or in general to the peroxide of a biatomic elementt , is less in accordance with what is at present considered true in chemistry , than to consider it as the peroxide of a monoequivalent element , analogous to peroxide of hydrogen .
It is remarkable that peroxide of hydrogen , in the solid stae or in solid compounds , must have almost as great an atomic heat ( for HO 2-3+4=6-3)-as those elements which obey DULONG and PETIT'S law , and especially as iodine , bromine , and chlorine , according to the direct and to the indirect determination of their atomic heat ; the same must be the , case for the analogous peroxides of such still unknown elements as have an atomic heat as great as that of hydrogen .
' As far as may be judged from its specific heat , chlorine may be such a peroxide ; but this consideration shows no necessity for assuming that it actually is so .
In a great number of cases the atomic heat of compounds gives more or less accurately a measure for the degree of complexity of their composition* .
And this is the case also with such compounds as are comparable in their chemical deportment to undecomposed bodies .
If cyanogen or ammonium had not been decomposed , or could not be so with the means at present offered by chemistry , the greater atomic heats of their compounds , compared with those of analogous chlorine or potassium compounds ( compare ? ?
96 ) , and of cyanogen and ammonium as compared with chlorine and potassium , would indicate the more complex nature of those so-called compound radicals .
The conclusion appears admissible that for the so-called elements the directly or indirectly ascertained atomic heats are a measure for the complexity of their composition .
Carbon and hydrogen , for example , if not themselves simple bodies , are more so than silicium or oxygen ; and still more complex compounds are the elements which are now considered as following DULONG and PETIT'S law ; with the restriction , however , that for these also the atomic heats may be more accurately determined and differences proved in them which justify similar conclusionst .
One might be tempted , by comparing atomic heats , to form an idea how the more complex of the present indecomposable bodies might be composed of more simple ones , just as such a comparison has been shown to be possible for chlorine ; but it is at once seen that to carry out such an attempt the atomic heats of the elements , certainty .
It may appear surprising , or even improbable , that so-called elements which can replace each other in compounds , as , for instance , hydrogen and the metals , or whichenter into compounds as isomorphous constituents , like silicium and tin , should possess unequal atomic heats and unequal complexity of composition .
But this is not more surprising than that indecomposable bodies , and those which can be proved to be compound , as , for example , hydrogen and hyponitric acid , or potassium and ammonium , should replace one another , preserving the chemical character of the compounds , and even be contained as corresponding constituents in isomorphous compounds .
I have here expressed suppositions in reference to the nature of the so-called elements which appear to me based on trustworthy conclusions from well-proved principles .
It is in the nature of the case that the certain basis of fact and of what can be empirically demonstrated must be left .
It must also not be forgotten that these conclusions only allow something to be supposed as to which of the present indecomposable bodies are more complex and which of simpler composition , and nothing as to the question what simpler substances may be contained in the more complex ones .
The consideration of the atomic heats may say something as to the structure of a compound atom , but in general gives no clue as to the qualitative nature of the simpler substances used in the construction of the more complex atoms .
But even if these suppositions are not free from uncertainty and imperfection , they appear worthy of attention in a subject which , for science , is still so much in darkness , as is the nature of the indecomposable bodies .
[ 71 ] III .
Investigations of the Specific Heat of Solid Bodies .
By HERMANN KOPP .
Communicated by T. GRAHAM , Esq. , F.RBS .
Received April 16 , -Read May 12 , 1864 .
I. Historical Introduction .
1 .
ABOUT the year 1780 it was distinctly proved that the same weights of different bodies require unequal quantities of heat to raise them through the same temperature , or on cooling through the same number of thermometric degrees , give out unequal quantities of heat .
It was recognized that for different bodies the unequal quantities of heat , by which the same weights of different bodies are " heated through the same range,.must be determined as special constants , and considered as characteristic of the individual bodies .
This newly discovered property of bodies WILKE designated as their specific heat , while CRAWFORD described it as the comparative heat , or as the capacity of bodies for heat .
I will not enter upon the earliest investigations of BLACK , IRVINE , CRAWFORD , and WILKE , with reference to which it may merely be mentioned that they depend essentially on the thermal action produced when bodies of different temperatures are mixed , and that IRVINE appears to have been the first to state definitely and correctly in what manner this thermal action ( that is , the temperature resulting from the mixture ) depends on the original temperature , the weights , and the specific heats of the bodies used for the mixture .
LAVOISIER and LALALCE soon introduced the use of the ice-calorimeter as a method for determining the specific heat of bodies ; and J. ; T. MAYER showed subsequently that this determination can be based on the observation of the times in which different bodies placed under comparable conditions cool to the same extent by radiation .
The knowledge of the specific heats of solid and liquid bodies gained during the last century , and in the first sixteen years of the present one , by these various methods , may be left unmentioned .
The individual determinations then made were not so accurate that they could be compared with the present ones , nor was any general conclusion drawn in reference to the specific heats of the various bodies .
2 .
DULONG and PETIT'S investigations , the publication of which commenced in 1818 , brought into the field more accurate determinations , and a general law .
The investigations of the relations between the specific heats of the elements and their atomic weights date from this time , and were afterwards followed by similar investigations into the relations of the specific heats of compound bodies to their composition .
In order to give a general view of the results of these investigations , it is desirable to present , for the elements mentioned in the sequel , a synopsis of the atomic weights assumed at different arid most trustworthy experimental determinations , without alteration of the bases for the adoption of these numbers .
The numerical relations presented in the above Table require , from the chemical point of view , no further explanation .
The relations of these numbers to the specific heat form the subject of the investigations which are presented in the sequel .
3 .
The experiments by which DULONG and PETIT * showed , in the case of mercury various solid metals , and glass , that the specific heat increases with increasing temperature , were made by the method of mixtures .
They determined at ordinary temperatures the specific heats of a greater number of elements by the method of cooling + .
They found that when the numbers in the first column in ? ?
2 corresponding to the elements Bi , Pb , Au , Pt , Sn , Zn , Cu , Ni , Fe , and S ( the Berzelian atomic weights ) are multiplied by the respective specific , heats of these bodies , approximately the same number is obtained ; and that approximately the same number is also obtained when X Ag , 1 Te , and ?
Co are multiplied by their corresponding specific heats .
They were of opinion that the atomic weights of the elements could and should be so selected that , when multiplied by the specific heats , they should give approximately the same number as product .
This observation and this view , which DULONG and PETIT stated in 1819 in the following manner , " The atoms of all simple bodies have all exactly the same capacity for heat , " have since that time been known as DULONG and PETIT 's Law : I shall not here dwell upon POTTER'S investigations on the specific heat of metals and on the validity of DULONG and PETIT'S law $ , but proceed directly to ; discuss NEUMANN'S investigations , which rank worthily by the side of those of DUIONG and PETIT .
4 .
In h Invion his Investigation on the specific heat of Minerals , " NEUMANN ( in 1831 ) first published ?
more accurate determinations of the specific heats of solid compounds .
He investigated a large number of such compounds , especially those occurring in nature , partly by the method of mixture , and partly by the method of cooling ; and he determined the sources of error in both these methods , and the corrections necessary to be introduced .
In a postscript to this paper , he mentioned that he continued the investigations with an apparatus which , compared with that he had previously used , promised far greater accuracy in the individual results , without needing tedious and troublesome reductions .
This apparatus , by means of which the specific heats of solid bodies , which may be heated in a closed space surrounded by steam , can be determined with great accuracy , he has not described II .
Of the general results of NEUMANN'S investigations , one must be particularly men* Annals de Chimie et de Physique , [ 2 ] vol. vii .
p. 142 .
t Ibid. vol. x. p. 395 .
+ Edinburgh Journal of Science , New Series , vol. v. p. 75 , and vol. vi .
p. 166 .
J. F. W. JOENsTON'S remarks , vol. v. p. 278 .
I only know these papers from BERZELIuS'S ' Jahresbericht , ' vol. xii .
p. 17 , and GEHLE'S 'Physicalisches Worterbuch , ' new edition , vol. x. part 1 , p. 805 et seq. ?
POGGENDORFF'S 'Annalen , ' vol. xxiii .
p. 1 .
11 PAPE ( POGGENDORFF'S 'Annalen , ' vol. cxx .
p. 337 ) has recently described this apparatus .
I have had no tioned , that a dimorphous substance has the same specific heat in its two conditions .
This he showed was the case with arragonite and calcite , and with iron pyrites and marcasite .
But the most important is the discovery that in analogous compounds the products of the atomic weights into the specific heats are approximately equal .
NEUMANN stated this last observation in the following manner:- " In bodies of analogous chemical composition the specific heats are inversely as the stochiometrical quantities , or , what is the same , stochiometrical quantities of bodies of analogous chemical composition have the same specific capacity for heat .
" NEUMANN adduced 8 carbonates , 4 sulphates , 4 sulphides ( Me S ) , 5 oxides ( Me 0 ) , and 3 oxides ( Me2 03 ) , as showing this regularity , which is to be denoted as NEUMANN'S law * .
5 .
Soon after the publication of NEUMANN'S researches in 1833 , AVOGADRO published ta Memoir on the Specific Heat of Solid and Liquid Bcdies .
" He there gave a number of determinations of the specific heat of solid bodies made by the method of mixture .
As far as can be ascertained by comparison with the most trustworthy of our newer determinations , these results are by no means so accurate as those of NEUMANN ; but they are far more accurate than those which had been obtained up to about 1830 , and many of them come very close to the best of our modern results .
It would be unjust to AVOGADRO'S determinations $ to judge them all by one case , in which he obtained a totally erroneous result ( for ice , by a modified method ) ; and by the circumstance that in a subsequent memoir ?
he gives specific heats for several elements as deduced from his experiments , which are decidedly incorrect J. AVOGADRO recognizes the validity of DULONG and PETIT'S law .
With reference to the specific heats of compound bodies , he considers that he had established , with tolerable probability , that for solid and liquid bodies the same regularity prevails which he had previously deduced for gases from DULONG'S experiments .
That is , ' " that the specific heat of the atom of a compound body is equal to the square root of the integral or fractional number expressing the atoms or parts of atoms which go to form the atom of the compound body such as it exists in the solid or liquid state , taking as unity the specific heat of the atom of a simple body in the samestate .
" He observes that there is a difficulty incidental to the application of this law to solid and liquid bodies , which is not met with in the case of gaseous bodies , in which the composition by atoms or by volumes is held to be directly given by opportunity of seeing NEUMANN'S memoir cited by PAPE , " Commentatio de emendenda formula per quam calores corporum specifici ex experimentis methodo mixtionis institutis computantur .
" Regiomonti , 1834 .
observation .
This difficulty consists in knowing what constitution is to be assigned to the body in question for the solid or liquid condition ; this constitution , from the conclusions derived from his theoretical considerations , would often be different from that which the body has in the state of gas or vapour .
His considerations led him to assume the atomic weights of many elements different from those which BERZELIUS had given : AVOGADRO described the atoms , to which the weights assumed by him refer , as thermal atoms .
6 .
R. HERMANN published in 1834 a memoir " On the Proportions in which Heat unites with the Chemical Elements and their Compounds , and on the Combining Weights considered as quotients of the capacity for Heat of Bodies into their Specific Gravities"* .
He gives there a great number of determinations of the specific heat of solid bodies ( of a few elements , but chiefly of compound bodies ) .
He made a few experiments in which he used LAVOISIER and LAPLACE'S calorimeter t ; but by far the greater number of determinations are made by the method of cooling $ .
Many of his results approach very closely to those which are at present considered accurate , but they are in so far untrustworthy that a considerable number among them are decidedly incorrect .
As for HERMANN'S theoretical results , it must be borne in mind that , regarding matter as he does , not from the point of view of the atomic but of the dynamical theory , he puts the idea of combination weights in the place of the idea of atomic weights .
The propositions which he endeavours to establish are the following .
The quotients obtained by dividing the ; specific gravities of the elements ?
in the solid state by their specific gravities in the gaseous state , are either equal or stand to each other in simple ratios ; they are 1 , 2 ... ... 15 times as much as a certain base .
The same is the case with the products of the specific gravities of the solid elements into their specific heats , that is , with their relative heat ; and the number indicating the multiple for a given element is the same for both the above relations .
It follows from this that the combining weights m of the elements are proportional to the quotients of their relative heats into their specific gravity in the solid condition ; that the , products of the specific heats and the combining weights for different elements are equal to a constant , and that from the known combining weight of an element its specific heat in the solid form may be calculated ( it is equal to 03 7 5 , where m is the combining weight of the substance in question referred to oxygen = 1 ) .
For several elements ( phosphorus , tellurium , cadmium , and silver for instance ) atomic weights are taken which differ from those of BERZELIUS .
In the case of the sulphides , the specific heats may be calculated from those of the constituents , assuming that the specific heats of the elements in these compounds are the same as in the free state .
The same holds good for several chlorides and for basic metallic oxides , if the specific heats of chlorineand of oxygen , as given by the above formula , are taken as basis .
But in acids a smaller specific heat must be taken for oxygen ( one half in several acids and null in phosphoric acid ) ; and there are even compounds ( cassiterite , e. g. , or arsenious acid ) , in which the same element is contained partly with the normal and partly with the modified specific heat* .
For oxygen salts it is to be assumed that both the acid and the base have the same specific heat as in the free state , and hence the specific heat of one constituent ( of the acid , for instance ) may be calculated , if that of the salt and that of the other constituent ( the base ) is known ; and it is also found that the specific heat of chromic acid in the neutral and in acid chromate of lead is the same .
This memoir of HERMANN'S did not become much known .
Unacquainted with it , other philosophers have subsequently developed independently similar opinions .
7 .
In 1835 RUDBERG described a methodt , which , by ascertaining the heat developed when salts are dissolved in water , in experiments in which the proportion of the salt to the water was constant , but the temperature of the salt varied , should give a means of at once determining the specific heat of the salt , and of the heat which was either absorbed or became free .
Yet the numbers which he obtained from his experiments for the specific heat of solid salts are undoubtedly erroneous .
DUMAS4 ( in 1838 ) discussed the possibility of determining the specific heat of organic bodies by the following process .
A platinum vessel containing the substance in question , along with a thermometer , is to be heated to 30 ?
or 40 ?
, and then brought into a vessel provided with a second thermometer , and containing water , the temperature being about 5 ?
or 6 ?
lower than that of the surrounding room .
When the temperature has risen to the same extent above that of the room , both thermometers are to be observed .
I know no determinations made by this method .
8 .
In 1840 REGNAULT commenced the publication of a series of important investigations on specific heat which he had made .
As they are generally known , I may be more brief in enumerating the contents of the individual publications .
In the first which he published , REGNAULT developed ?
the reasons which led him to prefer the method of mixture to other processes for determining the specific heats of solid bodies ; TiEG SPECIFIC HEAT OF SOLID BODIES .
B he described his mode of executing this method , and published the results obtained for a great number of elements .
In a second memoir * he gave the specific heats of several metallic alloys containing metals in simple atomic ratios , and of a great number of solid chemical compounds ; and he published comprehensive experiments on the specific heat of carbon in its different conditions .
The investigations announced in the first memoir t on the specific heat of organic compounds , as well as those promised in the second memoir4 on the specific heat of sulphur at different temperatures , have not to my knowledge been published .
But in a third memoir ?
REGNAULT has investigated the difference in the specific heats of certain metals according as they are hardened or soft , and also with reference to sulphur according as it is in the native crystallized form , or has solidified a longer or shorter time after being melted ; and he has more especially tried to impart greater certainty to the method of cooling .
In his subsequent investigations , however , he has only used the method of mixture as being the more certain .
These investigations 11 have given the specific heats of a large number of solid elements , and also of individual compounds .
By his investigations REGNAUJLT has removed some objections which seemed to affect DULONG and PETIT'S law , and has given a great number of new cases in which it applies .
He considers ?
this law to be universally valid , and discusses the reasons why for individual elements the specific heats found do not quite agree with the law , but only approximately .
In his view the atomic weight of an element is to be so taken that it agrees with DULONG and PETIT'S law .
He took the atomic weight of silver and of the alkaline metals half as great , and that of carbon twice as great as BERZELITUS had done .
Yet with regard to selecting , by means of the specific heat , from among the numbers which the chemical investigations of an element has given as admissible , that which is the correct one , REGNAULT does not always express himself decidedly .
In the case of carbon ** and of silicium ftt he mentions the possibility of their disagreement with DULONG and PETIT'S law .
He proved the validity of NEUMANN'S law for a ntimber of cases very considerably greater than that on which it had originally been .
based ; and he expressed it in a much more general form 4 .
" In all compounds of analogous atomic composition , and similar chemical constitution , the specific heats are approximately inversely proportional to the atomic weights .
REGNAULT designates the numbers agreeing with this law as thermal atomic weights .
Hle has either determined them directly from the numbers found for the specific heats of the elements in the free * Ann de Chim .
et de Phys. [ 3 ] vol. i. p. 129 .
t Ibid. [ 2 ] vol. lxxiii .
p. 71 .
+ Ibid. [ 3 ] vol. i. p. 205 . ?
Ibid. [ 3 ] vol. ix .
p. 322 .
11 Ibid. [ 3 ] vol. xxvi .
pp. 261 &amp; 268 ; vol. xxxviii .
p. 129 ; vol. xlvi .
p. 257 ; vol. Ixiii .
p. 5 .
Comptes Iendus , vol. lv .
p. 887 . ?
Ann. de Chim .
et de Phys. [ 2 ] vol. lxxiii .
p. 66 ; further , [ 3 ] vol. xxvi .
p. 261 , and vol. xlvi .
p. 257 .
state , applying DULONG and PETIT'S law , or indirectly by ascertaining the specific heat of solid compounds , assuming NEUMANN'S law ; or finally ( and only in a few cases ) , he hasdetermined them by means of their probable analogies .
These are the atomic weights given in the second column of the Table in ? ?
2 .
With regard to the relations of the specific heats of solid compounds to those of their constituents , REGNAULT has shown * that with metallic alloys , at a considerable distance from their melting-points , the specific heats may be calculated from those of their constituents in tolerable accordance with the experimental results , assuming that the specific heats of the metals are the same in the alloys as in the free state .
The investigation , whether for true chemical compounds there is a simple relation between their specific heats and those of their constituent elements , REGNAULT has reserved t till the conclusion of his experiments on the specific heats of gaseous bodies 4 .
To my knowledge he has published nothing for solid bodies .
But in 1862 , with reference to the relations which had been recognized between the specific heats and atomic weights of solid , simple or compound bodies , he spoke as follows t ? .
" It is true that these laws , in the case of solid bodies ; only apply approximately to simple bodies and those compounds of least complex constitution ; for all others it is impossible to pronounce anything in this respect .
" From some remarks of REGNAULT in reference to carbon [ 1 and silicium ?
he considers it possible , or probable with certain elements , that they have a differept specific heat in their compounds to that which they have in the free state .
9 .
In 1840 DE LA RIVE and MARCET published ** investigations on the specific heat of solid bodies .
They made their determinations by the method of cooling .
They found that , assuming BERZELIUS'S atomic weights , selenium , molybdenum , and wolfram fall under DULONG and PETIT'S law , which they consider as universally valid ; but that carbon forms an exception , and they consider it as probable that its true atomic weight has not yet been ascertained .
For several sulphides they found a greater specific heat than was calculated for them , assuming that their constituents have in them the same specific heat as in the free condition .
They think that for solid as well as for liquid and gaseous compounds the law governing the specific heat is still unknown .
A subsequent memoir by these physicists tt treated of the specific heat of carbon in its various conditions .
10 .
In 1840 ++ H. SCHRODER made an investigation as to what volumes are to be assigned to the constituents of solid and liquid compounds when contained in those compounds .
In his memoirs on the subject , he expressed the view that the specific heat of compounds depends on the specific heats of the constituents in that particular * Ann. de Chim .
et de Phys. [ 3 ] vol. i. p. 183 .
t Ibid. p. 132 .
+ REGNAULT has made known the results of these experiments in 1853 by a preliminary account in the Comptes Rendus , vol. xxxvi .
p. 676 , and more completely in 1862 in his ' Relation des experiences pour determiner les lois et les donnees physiques necessaires au calcul des machines a feu , ' vol. ii .
p. 3 . ?
Relation , &amp;c. vol. ii .
p. 289 .
1 Ann. de Chim .
et de Phys. [ 3 ] vol. i. p. 205 .
1 Ibid. [ 3 ] vol. Ixiii .
p 31 .
state of condensation in which they are contained in the compounds in question .
In 1841 * , reasoning from the results of REGNAULT'S experiments , he endeavoured to show that the atomic heat ( that is the product of the atomic weight into the specific heat ) of a compound is equal to the sum of the atomic heats for the states of condensation in which the elements are contained in the compound , and to ascertain what atomic heats are to be assigned to certain elements in certain compounds .
On the assumption that the atomic heat of metals in compounds is as great as in the free state , he endeavoured to determine the atomic heat of oxygen , sulphur , &amp;c. in certain compounds of these elements with the metals ; he came to the conclusion that an element ( sulphur for instance ) may in some compounds have an atomic heat different from that which it has in the free state ; and the same element ( sulphur or oxygen for instance ) may have different atomic heats in different compounds ; but the changes in the atomic heat .
of .
an .
element always ensue in simple ratios .
I cannot here adduce the individual results , which he obtained when he inferred the atomic heat of an element in a compound by subtracting from the atomic heat of the compound the atomic heat of the other elements in it , which he had calculated either from direct determinations of their specific heat , or from previous considerations .
The essential part of SCHR6DER'S conception is that in this manner the atomic heat of a body , as a constituent of a compound , , may be indirectly determined ; and the result is that the atomic heat , at any rate of some elements in compounds , is different from what it is in the free state , and may be different in different compounds , and that the changes are in simple ratios .
SCHRODER considered also that there was probably a connexion between these changes and those of the volumes of the elements , without , however , stating how from the one change the other might be deduced .
11 .
L. GMELIN ( in 1843 ) considered it as inadmissible , from the chemical point of view , .
to assign throughout such atomic weights to the elements as to make them agree with DULONG and PETIT'S law .
Certain exceptions must be admitted .
Comparing the specific heats of oxygen , hydrogen , and nitrogen for the gaseous state with the specific heats of other elements in the solid state , he came to the conclusion that if the numbers given in ? ?
2 as the equivalents ordinarily assumed be taken as atomic weights , the atomic heat of hydrogen , of nitrogen , and by far the greater number of the elements is equal to about 32 ; several of them twice as great , that of oxygen one-half , that of carbon ( as diamond ) one-fourth as great .
With reference to the dependence of the atomic heats of the compounds on those of the elements , GMELIN expressed the opinion + that in general the elements on entering into compounds retain the atomic heats they have in the free state , but for individual elements , especially for oxygen and carbon , it must be assumed that their atomic heat changes in simple ratios with the compounds , into which they enter .
* POGGENDORFF'S ' Annalen , ' vol. lii .
p. 269 .
t L. GMELIN'S 'Handbuch der Chemie , ' 4th ed. vol. i. p. 217 .
was also of opinion * that the specific heats of the elements remain unchanged when they enter into chemical compounds .
In 1848 he stated as a general proposition ; C " The quantity of heat necessary to raise the temperature of the atomic weight of a body through 1 ?
is equal to the sum of the quantities of heat necessary to raise the temperature of the atoms , and fractions of atoms , through 1 ?
" .
If A is the atomic weight and C the specific heat of a compound , :a , , a2 , a3 ... .
the atoic weights t , and cl , c2 , ... .
the specific heats of the elements contained in it , and ni , n2 , n3 ... the numbers which express how many atoms of each element are contained in an atom of the compound , then AC-naic , +n2a2c2+ n3a33 ** As a proof of this law , he compared the calculated values of AC of several compounds ( metallic iodides and sulphides ) and alloys with the observed values , taking REGNAUXT'S determinations of the specific heats of the elements and of the compounds .
It follows , further , from that proposition , that if the formula and the values for several compounds are compared with each other , there must be the same differences of the values AC for the same differences of formulae .
WCTESTYN showed by a number of examples that this is so approximately .
By means of this law , the product of the specific heat and the atomic weight for one constituent of a compound may be found , if this is known for the compound and the other constituents .
WaESTYN deduced in this way the product for oxygen ( by subtracting from the product for different metallic oxides that found for the metals , and from chlorate of potass that for chloride of potassium ) to be 2A4 to 2-1 ( 0 , = 8 ) , and for chlorine 3-0 to 3'5 ( C1.C=17 75)# WVESTYN finally expressed a doubt whether NEUMANN'S law is universally applicable .
He laid stress on the circumstance that when two elements give different products , the difference is also met with in the products for their analogous compounds ; and , for instance , the greater products which mercury and bismuth have in comparison with other elements , are also met with in the compounds of these metals .
13 .
GARNIER ( in 1852 ) developed the view$ , that not only in the case of elements are the atomic weights A ?
inversely proportional to the specific heats C , but that the same is the case with water || and solid compounds in whose atom n elementary atoms are A contained , if the so-called mean atomic weight be compared with the specific heat C ; for elements Ax C=3 , and for compound bodies X C= 3 ( if 0=8 ) .
He endeavoured to prove this from REGNAULT'S determinations of specific heats .
From the latter equation he calculated the specific heat for several compounds .
In the case of the basic oxides , sulphides , chlorides , bromides , and iodides , his calculated results agree tolerably with the observed ones ; this is less the case with metallic acids and oxygen salts , for which calculation mostly gives tesullts far too large .
GARNIER* drew , further , &lt; from the above proposition the conclusion , that the atomic weight of hydrogen , chlorine , &amp;c. must in fact be taken only half as great as the equivalent weight ; for only by assuming this smaller atomic weight is the mean atomic weight such that its product with the specific heat is near S. In 1862 BANCALARTt repeated that the specific heat of an atom of a compound body ( that is , its atomic heat ) is equal to the sum of the specific heats of the individual constituent simple atoms , and showed , from a series of examples ( oxides , chlorides , sulphates , and nitrates ) , that , according to that proposition , the atomic heats of many compounds may be calculated in tolerable approximation with those derived from REGNAULT'S experimental investigations , if , for the elements which he investigated , the atomic heats derived from his determinations be taken as a basis , that is , for oxygen ( 0=8 ) the atomic heat 1-89 ; for chlorine ( C1=17'75 ) 3-21 ; for nitrogen ( N=7 ) 3-11 .
CANNIZARO ( in 1858+ ) has used the proposition , that , in the sense above taken , uniAC versally C-=a constant , for the purpose of ascertaining the value of n for the atomic weight of different compounds , and therewith ascertaining the atomic weight of elements which are contained in these compounds .
14 .
Besides those of REGNAULT , but few experimental determinations of the specific heats of solid bodies have been published .
BED ?
and BYSTR6M 1| have published investigations on the specific heat of several metals at different temperatures ?
" : both sets of experiments were made by the method of mixtures .
From the year 1845 , PERSON * , in his investigations on the specific heat of ice , then on the latent heats of fusion , and their relations to the specific heats in the solid and liquid condition , has determined the specific heat for several solid substances , especially also for some hydrated salts .
He worked more especially by the method of mixture .
He observed ttf , in the case of these * Comptes Rendus , vol. xxxvii .
p. 130 .
t An abstract from Memory della Accademia dell Scienze di Torino , [ 2 ] vol. xiii .
p. 287 , in the Archives des Sciences Physiques et Nlaturelles , vol. xxii .
p. 81 .
I only know the contents of this memoir from this abstract .
+ Il Nuovo Cimento , vol. vii .
p. 321 .
PIAZZA also gives a statement of this specutlation in his pamphlet , 'Formole atomistiche et typi chimici , ' 1863 .
I only know this from a notice in the Bulletin de la Societe Chimique de Paris , 1863 . ?
An abstract from the Bulletin de rAcademie des Sciences de Belgique , vol. xxii .
p. 473 , and the Memoires Couronnes par I'Academie de Belgique , vol. xx:vii .
, appeared in the Bericht iiber die Fortschritte der Physik im Jahre 1855 , dargestent von der physicalisehen Gesellschaft zu Berlin , p. 379 . !
| Abstract from the Oversigt of Stockholm Vetenskaps-Akademiens Forhandlingar , 1860 , in the same Jahresbericht , 1800 , p. 369 .
I To the experiments of DtONGo and PETIT on this subject , mentioned in ? ?
3 , POUILLET'S determinations of the specific heat of platinum at different temperatures must be added ( Comptes Rendusy Vol. ii .
p. 782 ) .

