]\gt ; The Capacity for Heat of Metals at Different Temperatures .
549 nitrogen sulphide .
Stannic chloride and titanium tetrachloride also yield solid products .
In the latter case nitrogen was proved to be present .
( 5 ) All organic compounds tried , except carbon tetrachloride , yield hydrocyanic acid freely , but not cyanogen , as was proved by chemical tests .
When chlorine is present , chloride is formed .
Benzene yields ( almost certainly ) cyanobenzene .
( 6 ) The intensity of the cyanogen spectrum with organic is no index of the quantity of hydrocyanic acid being formed .
Preponderance of the red cyanogen bands is associated with cyanogen chloride or bromide .
On a general view of the evidence , there does not appear to be any definite connection between the development of spectra by active nitrogen and the chemical actions in progress .
I take this opportunity of thanking my colleagues , Prof. H. B. Baker and Dr. M. O. Forster , for constant help and advice .
The Capacity for Heat of Metals at Different Temperatures .
By E. H. GRIFFITHS , Sc. D. , F.R.S. , and EZER GRIFFITHS , B.Sc. , Fellow of the University of Wales .
( Received April l , \mdash ; Read May 1 , 1913 .
) ( Abstract .
) The object of this investigation was the determination , with the closest approach to accuracy , of the specific heat of certain metals , and the changes therein caused by changes in temperature .
The work , including the necessary preliminary standardisations , has occupied our attention during the last two years and we feel unable , in this abstract , to describe with any clearness the various precautions adopted , or the somewhat elaborate process of reducing the results .
We , therefore , here content ourselves with a brief indication of the methods of experiment , a statement of the results , and a short discussion of certain suggestive features which they present .
I. The method of experiment is briefly indicated in the following numbered paragraphs :\mdash ; VOL. LXXXVIIL\mdash ; A. Griffiths aGriffiths .
dependent upon any assumption concerning the capacity for heat of bodies !
: other than those under consideration .
: 2 .
The substances were raised across a given temperature through very small ranges of temperature ( extreme limit of range , about 3 .
These temperature changes were measured by means of differential platinum thermometers , for which purpose these instruments are admirably adapted .
4 .
masses of the substances were used , ranging from 1 to 4 kgrm .
5 .
The apparatus was constructed with all the parts duplicated .
The metals examined were suspended by quartz tubes in similar air-tight brass cases , which were placed side by side in a large tank containing rapidly stirred water or oil .
This tank was electrically co1ltrolled with great constancy at any given temperature One of the metal blocks remained at the tank temperature throughoub an experiment , while the other , having been previously cooled below , was raised to a somewhat similar temperature above it by a supply of heat electrically developed in the centre of the block , the difference in the temperature between the two blocks being determined at regular intervals by means of the differential platinum thermometers .
Any changes in the surrounding conditions would , therefore , affect both blocks equally , and hence , by measuring the difference of temperature only , possible causes of error were eliminated .
6 .
equation connecting the various quantities is MS where total mass and its specific heat , the initial temperature and the final temperature , the potential difference at the extremities of the coil of resistance , and the number of thermal units lost , or gained , during time from sources other than the electrical supply .
In these experiments the values of and were so arranged that was in every case small or negligible , and , if necessary , could be ffitimated with sufficient accuracy .
7 .
With two exceptions ( Cu and Fe ) the samples of metals used were supplied by Messrs. Johnson and Matthey , to whom we wish to express our , sincere thanks for the trouble they have taken in the matter .
Each sample was accompanied by their certificate egarding its purity .
*For a full discussion of the reasons for selecting this value of see p. 110 of the Thermal Measurement of Energy , ' by E. H. ( .
Univ. Press ) .
1913 .
] for of Metals at fferent Ter 5 51 An analysis of the iron was supplied by the manufacturers , and Mr. C. T. Heycock has very kindly made an analysis of the copper .
8 .
Experiments on identical samples at the same temperature were repeated under very varied conditions .
Two separate methods of experiment , involving different data and methods of reduction , were employed .
Three different sets of platinum thermometers were used .
The rate of heat supply was varied in the ratio of 9 : 1 .
The determination of at a given temperature , with a particular sample , was in several cases repeated after the lapse of some months .
The quartz tubes and their cover were replaced by others of diflerent masses , etc. We were thus enabled to ascertain causes of error which would otherwise have remained undetected .
9 .
The results of our observations have been deduced from the actual experimental numbers and in no case from " " smoothed culves.\ldquo ; The most serious difficulty presented by this method of experiment is that of determining the mean temperature of the block of metal when its temperature is altering .
Temperature gradients must necessarily exist , since equalisation of temperature by stirring is an impossibility .
The manner in which this ifficulty was surmounted is described in full in the original paper .
When embarking on this investigation we proposed to extend our of temperature to the lowest point obtainable by means of liquid air , limiting the inquiry to a study of two or three metals only .
Owing , however , to delay on the part of the contractors in the delivery of the liquid-air plant , we were compelled to postpone that portion of our investigations dealing with temperatures below C. to a later date , and therefore we enlarged the scope of our inquiry so as to include the metals , namely : aluminium , iron , copper , zinc , silver , cadmium , tin , and lead .
As the data already accumulated concerning the capacity for heat of these metaJs over the C. to 10 may be useful to other observers , we have seen no reason for postponing the publication of the work already completed .
II .
Statement of In the following tables we have , in each case , inserted in Column IV the " " probable error expressed in percentages , as the resulting numbers are of a nature suitable for treatment by the " " method of least squares and this appeared to be the most concise way of indicating the extent of the discrepancies in the results .
This " " probable error\ldquo ; does not , of course , include persistent errors or those present in determinations of the various standards employed .
We have , however , no reason to suppose that 552 Dr. E. H. ffiths and Mr. E. Griffiths .
[ Apr. 1 , such errors are of sufficient magnitude to have any appreciable effect upon our conclusions .
In order to express the relation between and , many forms of equations were tried , but it was found that the mean path of the results was most closely expressed in all cases by a parabola .
In Column we give the differences ( expressed in percentages ) between our values as found by experiment and those deduced from the parabola given at the foot of each table .
We also give the previous treatment of the metal and its analysis .
With copper an unusually large number of experiments at C. were performed .
Two methods were employed which involved different data .
The close correspondence between the resulting numbers convinced us of the validity of both methods , and therefore , in our later work , we contented ourselves with the adoption of the more convenient , i.e. that which we termed the " " method of intersection The capacity for heat of the copper case appeared as a small correction in our determinations of the specific heat of all the remaining metals .
For the above reasons we made no less than 35 determinations of its specific heat at C. , and these were performed on different dates and under varied conditions .
In two cases , where a note of interrogation is placed after a group number , it is an indication that we regard the result as less satisfactory than usual .
* Although little weight should be attached to values thus indicated , we do not feel that we have evidence sufficient to justify us in discarding them .
The " " group numbers\ldquo ; ( Column II ) do not merely indicate repetition of the same experiment , for the rate of supply of heat differed in the individual members of the group .
For example , in Table I , Column II , the first group of 12 includes experiments performed with potential differences varying from 3 up to 9 Weston cells , hence the rate of heat supply was , in this group , changed in all the following proportions : 9 : 16 : 25 : 36 : 49 : 64 : 81 .
In no group in any of the tables are there less than three such changes .
Column indicates the close correspondence of results obtained under such conditions .
With the exception of Cu and Fe , Messrs. Johnson and Matthey state that the previous history of the blocks was as follows:\mdash ; " " The cylinders in every instance were cast and then allowed to cool , subsequently being turned in a lathe .
They were not annealed The weight of metal as given at the head of each table is only a rough We have some doubts as to the constancy of the bath temperature during these two groups .
1913 .
] for Heat of Metals at Different .
553 approximation , as the mass was , in some cases , slightly altered series of experiments .
The copper was electrolytically deposited .
Table I.\mdash ; Copper .
Weight .
Density Mr. C. T. Heycock writes as follows:\mdash ; " " Cu per cent. Remaining per cent. consists of traces Pb , Fe , and a very little .
You will be correct in stating that it is of high purity During the above experiments at , three different pairs of thermometers were employed , and two different covers to the copper case .
The quartz tubes holding the block were also altered .
The frequent repetition of group experiments was considered necessary in order to ascertain the effect of such changes .
With reference to the sample of aluminium employed ( Table II ) Messrs. Johnson and Matthey state : " " Aluminium we have reason to believe to be exceptionally pure , say per cent. , with a trace of iron With the exception of the first group of 3 , these experiments were extremely satisfactory , so much so that the fifth figure appears to have some real gnificance .
The perfect agreement of the experimental and curve values is very noticeable .
554 Dr. E. H. Griffitf Weighi .
1913 .
] for Heat of at Diferent Temperatures .
This specimen was presented to us by the American Rolling Mm Company , Middletown , Ohio , U.S.A. , who state : 'Material rolled from an ingot into a billet ( 4 in .
by 4 in .
) on 'blooming mill , ' forged into round section at blacksmith 's shop .
Same had no further annealing nor additional treatment other than when rolled and forged The specimen was turned down to size in the laboratory workshop .
Analysis supplied by the manufacturers:\mdash ; Iron .
Mn .
Sitrace . .
Cu Table Zinc .
Weight .
Density .
*Where the " " value adopted\ldquo ; is not the mean of the numbers in Col. III in this , or other , tables , the reasons are dicated in the original paper .
Messrs. Johnson and Matthey state " " approximately per cent. zinc The agreement between the results , on repetition at the same temperature , was less satisfactory than usual ( see Column III ) , the extreme difference from the adopted value at C. being per cent. 556 Dr. E. H. ffiths and Mr. .
Griffiths .
[ Apr. ] , Table Silver .
Weight .
Density .
Messrs. Johnson and Matthey state : " " Better than fine Table Cadmium .
Weight .
Density .
Messrs. Johnson and Matthey state : " " Fully per cent. pure , with very slight traces of iron and zinc 1913 .
] for Heat of Metals at Different Table Weight .
Density experiments igroupNIoIof.roup .Messrs .
Johnson and Matthey state : " " Probably analyse to per cent. with trifling quantities of arsenic , lead , and iron Table VIII .
Lead .
Weight .
Density .
Messrs. Johnson and Matthey state : " " Approximate to per cent. , with inappreciable traces of arsenic and bisnluth The numbers in Column indicate that the agreeme1lt between the individual membsrs of a group in Table VIII is distinctly inferior to that attained with metals .
This is probably effect of the low conductiviby of lead and the consequent steepness of the thermal radients within the cylinders .
558 Dr. E. H. Griffiths and Mr. E. Griffiths .
[ Apr. 1 , III .
An inspection of these results will show that the curvature in the case of Al and Fe is far more marked than in the metals .
The curves of Sn and present certain interesting features .
The specific heat of Sn at appears , for reasons given elsewhere , to be exceptionally high .
This may in some way be connected with the fact that Sn at temperatures not much below readily assumes the " " grey powder\ldquo ; form and it is possible , therefore , that at its physical condition is an exceptional one .
The curvature in the case of shows distinct alteration as the temperature approaches , the rate of increase being much diminished .
In consequence , in this case , we pushed our inquiry to a temperature exceeding 12 C. The results confirmed the indications observed at somewhat lower te , nlperatures .
In this connection it should be remezubered that becomes malleable about 12 C. and we have some indication of a physical cause for its abnormal behaviour .
The range of atomic weights covered by the metals enumerated above is from ( A1 ) to ( Pb ) .
At the conclusion of our experiments , we proceeded to plot our values of at C. as ordinates and the atomic weights as abscissa .
The resulting points lie very closely on the curve [ deduced from the values of Al , the mean of Cu and , and Pb ] which is represented by where atomic weight .
The value of for Sn , however , exceeds the curve value by as much as per cent. ( i.e. , as against We have endeavoured to ascertain how nearly the values of obtained from this curve are in harmony with the experimental results of observers in the case of elements not examined by us .
In a large number , no satisfactory information is procurable and , where information exists , it is difficult to estimate its true value .
Again , few observers have used a ange of temperature including C. , and , in such cases , we have had to reduce their results to by means of the values given them at other temperatures , assuming the changes to be of a linear order .
In our full paper we have stated , in every case , the authority and the data by which we ascertained the most probable values at C. The following table shows the results of the ation : Column I gives the percentage difference between the calculated and the experimental values ( a positive sign indicating the latter as the greater ) in those cases where the difference does not exceed 3 per cent. In the case of 1913 .
] for Heat of Metals at Different Temperatures .
all gases , however , the experimental values have been doubled before making the comparison .
Column II .
Elements whose differences lie between 3 and 16 per cent. Of , however , we do not consider the values of , Na , and to be sufficiently established to warrant any conclusions , and we may state that , in order to test this point , we are now engaged in the determination of the specific heat of Na .
Column III .
* Elements whose experimental values differ so greatly from the calculated as to exclude the possibility of agreement .
A couple of curious coincidences , however , present themselves .
The calculated value of is almost exactly four times that of the diamond .
The mean value for amorphous is very closely half the calculated .
Table IX .
We are aware of the serious difficulties in the way of accepting any such relation as that given by the above equation between the atomic weights and specific heats at an arbitrary temperature such as C. It is , however , possible that the large majority of elements are in a stable condition at that temperature , and , if such is the case , some definite connection may exist between their atomic weights and specific heats .
However this may be , one thing is evident , , that the curve gives , throughout the whole range of atomic , valuss of in the case Column III the actual numbers , insead of percentage differences , are given .
The expression " " Atomic heat \ldquo ; is evidently an alternative manner of expressing the same relation .
560 The for Heat of at Different Temperatures .
of which , in a very majority of cases , are within 2 per cent. of the most probable values .
* It has already been pointed out that the curvature of the curve in the case of Al and Fe differs markedly from that of the remaining metals on our list .
If we take those six remaining metals , and , from the parabolic formulae given supra , deduce the point of intersection with the ordinate at , and , assuming the values of , find their atomic heats at absolute zero , we obtain as the mean of our results the number , their differences from the mean surprisingly small , when we consider the effect of any error in the values of the coefficients over the to 10 C. If then , assuming the values of at given by , and the experimental values of at and l00o , we construct the resulting parabolas and ascertain the differences between the experimental and the calculated values at the various points ally determined by us , we find that in no instance does the difference exceed per cent. , and that it is , in cases , much less .
The remarkable approximation between the hypothetical value of the atomic heat at C. of a body with atomic weight 1 and the likewise hypothetical value of the atomic heat of the group of metals at absolute zero is probably a coincidence , but may possibly be of real significance .
If we assume the value of Dulong and Petit 's constant as , the deduced values for the elements in Column I would , in some cases , differ from the experimental ones by as much as 30 per cent.

