On the Thermo-electric Series .
By AUGUSTUS MATTHIESSEN , Ph.-D .
Connnunicated by CHARLES WHEATSTONE , Esq. Received November 20 , 1857 , -Read January 14 , 1858 .
BEING enabled by the method described in the Philosophical Magazine ( February 1857 ) to obtain wires of the metals of the alkalies and alkaline earths , I have determined their places , together with most of those of the other metals and some alloys , in the thermoelectric series .
The alloys which were experimented with , are those described by ROLLMAN* as giving with other metals stronger thermoelectric currents than those of bismuth or antimony under the same circumstances .
If A , B , C are different metals , and ( A , B ) , ( B , C ) , ( C , A ) the electromotive powers of thermo-elements formed out of each two of these metals , whose alternate soldering-points are at two different temperatures , then ( A , B ) +(B , C ) + ( C , A ) = 0 , and thereforet ( A , B)=a-b , ( B , C)=6-c , ( C , A)= c-a , where the values of a , b , c not only depend on the two temperatures , but also on the nature of each of the metals A , B , and C. As the " differences " of the same constitute the electromotive powers , the value for any one of these metals may be put = 0 .
If the temperatures of the soldering-points of a thermo-element only vary slightly , the electromotive power may be said to be proportional to the difference of the two temperatures , and under the same conditions the quantities a , b , c are also proportional to the difference of the temperatures , and their ratios to each other therefore independent of the same .
If now the value of a second metal relative to the above-named value of the first be taken -1 the values of the others in relation to these can be deduced , and only depend on the nature of each metal .
These values I will call the Thermo-electric Numbers of the metals .
The results obtained are given in Table I. , where the thermoelectric number of chemically pure silver is taken 0 , and that of a certain commercial sort of copper =1 .
TABLE I. Bismuth ( commercial , pressed wire).+ 35-81 Bismuth ( pure , pressed wire ) .
+ 32-91 Alloy of 32 parts of bismuth and I part of antimony ( cast).+ 29 06 Bismuth ( pure , cast).+ 24-96 Bismuth ( crystal , axial ) .
+24-59 Bismuth ( crystal , equatorial).+ 17*17 Cobalt No. 1 ( a pressed specimen prepared by Professor DUFLOS , and out of the Collection of the Heidelberg Chemical Laboratory ) .
+ 8-977 Potassium ( the same as used for the determination of its electric conducting powers for different temperatures).+ 5 492 Argentine ( wire of commerce , hard drawn).+ 5-240 Nickel ( commercial , free from cobalt , but containing iron , &amp;c. ) ( from Messrs. EVANS and ATKIN , Birmingham).+ 5020 Cobalt No. 2 ( from the Collection of the Heidelberg Laboratory ) + 3-748 Palladium ( wire , hard drawn , from DESMOUTIS , CHAPUIS and Co. of Paris ) + 3-560 Sodium ( the same as used for the determination of its electric conducting powers for different temperatures).+ 3-094 Quicksilver ( pure , fused in a glass tube).+ 2'524 Aluminium ( from ROUSSEAU Fre'res of Paris , wire , hard drawn , analysed by Dr. G. C. CALDWELL , and found to contain Si 2-34 , Fe 5-89 , and Al 91P77 ) +1 283 Magnesium ( wire , pressed).+ 1-175 Lead ( pure , pressed wire).+ 1-029 Tin ( pure , pressed wire ) .
The difference of the electromotive powers of the elements was here so small that it could not be determined ; with the galvanometer , a copper-tin element , when the soldered points had 100 ?
C. difference , gave no appreciable deflection ; also the central experiments show as well how very little the electromotive power of a silver-tin thermo-element varies from a silver-copper one ... + 1000 Copper No. 1 ( wire of commerce , annealed , containing small quantities of zinc , tin , lead and nickel)..+ 1-000 Copper No. 2 ( wire of commerce , annealed).+ 0922 Platinum ( wire from DESMOUTIS , CHAPUIS and Co. of Paris ) .
+ 0723 Gold ( wire , hard drawn , purified by Dr. C. MEYBOOM).+ 0613 Iridium ( from the Collection of the Heidelberg Chemical Laboratory ) + 0163 Antimony ( wire , pressed specimens , purified by Dr. W. P. DEXTER and Dr. G. C. CALDWELL).+ 0036 Silver ( pure , drawn , hard ) .
0 000 Gas-coke ( from the Heidelberg Gas-Manufactory , the hard mass remaining in the retorts ) ... ... .oeo57 Zinc ( pure , pressed ) ... ... 0.208 Copper ( galvanoplastically precipitated ) ... . .
0.244 Cadmium ( a pure slip of foil from Professor B6TTGER ) 0 332 Antimony ( commercial , pressed wire). .
1 897 Strontium ( pressed wire). .
2028 Lithium ( pressed wire). .
3768 Arsenic ( a piece , pure ) ... ... e e..ee ee e-3e828 Calcium ( pressed wire ) ... .
4-260 Iron ( pianoforte wire No. 4 ) .
5 218 Antimony ( axial ) ... .6-965 Antimony ( equatorial ) ... .
9.435 Red phosphorus ( from Professor SCHR6TTER , from the Collection of the Heidelberg Chemical Laboratory ) ... 9-600 Antimony ( purified as above , cast ) ... .
9-871 An alloy of 12 parts of bismuth , 1 part of tin , cast ... .131370 An alloy of 2 parts of antimony and 1 part of zinc , cast ... .
-22 700 Tellurium ( from M. ALEXANDER LOEWE of Vienna , purified by M. HOLTZMANN , cast ) ... -179 80 Selenium ( from the Collection of the Heidelberg Chemical Laboratory. .
-290-00 The electromotive power of a copper No. 1 , Argentine thermo-elements,4whose soldering-points had the temperatures 7 ?
and 25 ?
, was found by Dr. WILD to be equal to 18-4 .
106 in the standard given by Professor WEBER , where the unit of length is 1 millimetre , the mass a milligram , and the time a second ; and that of the same thermo-element , where the difference of temperature at the soldering-points was 100 ?
C. , was found equal to 1 l108 compared with that of a IDANIELL'S element , the electromotive power of which was taken equal 1000* .
The above series agrees with that found by SEEBECK , with the exception of arsenic , which stands below iron according to his experiments ; this however is easily explained , as different sorts of iron give with each other strong thermoelectric currents .
The manner in which the experiments were carried out is the following:-Two thermo-elements , whose warm and cold soldering-points had the same temperatures , were compared with each other ; these formed a circuit with the coil of a galvanometer , which surrounded a magnet rod of about a pound weight , to which was fastened a piece of looking-glass , thereby allowing the deflections of the magnet to be observed at a distance by means of a telescope and scale , in the same way as observations are made with the magnetometer .
The commutators were brought into the circuit ; the one changed the direction of the current in the wire of the multiplier , and the other allowed the currents of the thermo-elements either to pass in a direction so that they strengthened , or so that they opposed each other .
Out of the four observed stationary deflections of the magnet which correspond with the four different possible positions of the commutators , the ratio of the electromotive powers of the two thermo-elements may easily be calculated ; for if so , S , S3 , 84 be the four observed deflections , and if sI and 84 be the combination of the two warm solderingpoints of the two elements , and s2 and s8 that of the warm soldering-points of the one with the cold one of the other , that ratio is ( S2 S3 ) + ( S S4 ) ( S 2S3 ) -(SI -S4 ) ' All the solderings of the different metals were placed in two cubical zinc boxes ( of about 6 inches the side ) ; these had double sides , the space between them being filled with ashes .
The one box was filled with warm water , and the other with water having the temperature of the room ; the wires were introduced partly through holes in the covers and partly through holes in the sides , which were made water-tight with indiarubber tubing .
In order to prevent any conducting of the electric current through the water , the wires were covered with thin india-rubber tubes , and by means of stirrers a uniformity of temperature was kept up throughout each of the boxes .
In consequence of their size and their double sides , the temperature hardly varied during the time an observation was being made .
If the electromotive powers of a thermo-element were in exact ratio with the difference of the temperatures , the knowledge of the same would not be necessary with the arrangement described ; but as this is not exactly the case , the temperatures were observed by means of thermometers placed in the boxes through the covers .
Table II .
, the arrangement of which will be explained immediately , contains the results of the observations made in the manner above described .
TABLE II .
Silver , Bismuth ( commercial wires)-5 Silver , Copper .
Mean .
tl=7-6 t2-24O0 V=7K163 ) t1=7-6 t2=23O0 V=7-164 7'162 t , =7-6 t2=22-4 V=7.1593 Silver , Bismuth ( pure wires)-5 Silver , Copper .
tj=4-2 t2=23.7 V=6 585 ) t , =42 t2=22-6 v=6-586 S6582 t1= 42 t2=21-8 V=6.575J Silver , Potassium---2 Silver , Copper .
t , =4-0 t2=29-3 V=2748 ) t , =4.0 t2=27.3 V=2-749 S2-746 to=4-0 t2=25-9 v=2 742J Silver , Argentine-2 Silver , Copper .
t1=5X6 t2=32K1 v=2 620 ) tt -5X6 t2=31 0 v=2 621 2620 t-=5-6 t2=300 V=2.620J Silver , Palladium-2 Silver , Copper .
ti=4X4 t2=27-7 V=1783 ) to=4 4 t2=27X2 V=1-775 1 780 t4-44 Et2=260 V=1 782J Silver , Sodium-2 Silver , Copper .
tl=4-8 t2=28-9 v=1-549 ' t1=4X8 t2=27O0 V=1-546 5 t1=4X8 t2 2601 V=1550 ( Continued .
) Silver , Quicksilver-2 Silver , Copper .
Mean .
t1 = 7'2 t2=2803 v=1'265 ) tl= 7-2 t2=27-8 v=1-262 1-262 t1 = 7'2 t2=27-0 v=1h259J 2 Silver , Copper-Silver , Aluminium .
tI= 50 t2t-280 v=15591 tl= 5-0 t2=26-9 v=1=558 1I558 t1 = 5-0 t2=26-0 v=1.558J 2 Silver , Copper-Silver , Magnesium .
t1= 6-8 t2= 301 v=1-698 ) t1= 6-8 t2=29-8 v=1-704 &gt; 1-702 t , = 6-8 t2=27-5 v=1.703J Silver , Lead-Silver , Copper .
t1= 9-0 t2=29-2 v=1O031 tl= 9-0 t2=28-8 v=1-031 1-029 t , = 9-0 t2=27-4 v=1-025J Silver , Copper-Silver , Copper ( No. 2 ) .
t1=25-2 t2=4249 v=1-084 ) t1=25.2 t2=42-2 v=1083 1-084 t1 25.2 t2=43-9 v=1-085j Silver , Copper-Silver , Platinum .
t , =26 2 t2=44 0 v=1-385 ) t1=262 t2=4358 v=1h388 1 383 t1=26 2 t2=42-0 v=1.376j 2 Silver .
, Gold-Silver , Copper .
t1= 4.5 t2=32A4 v=1-224 ' tj= 45 t2=31-3 v=1-222 t1= 45 t2=30 0 v=1-230 1226 t1= 45 t2=29 0 v=1-229 2 Silver , Copper-2 Zinc , Silver .
t1= 70 t2= 304 V=4 822 ) t= 70 t2 =29[1 V=4800 4810 t= 70 t2=26-6 V=4 8083 Silver , Copper-Cadmium , Silver .
t1= 5.4 t2=26-7 V=3O0024 t 5.4 t2=25-8 V=3 008j3 007 t_= 5.4 t2=25 2 V=3'O11 ( Continued .
) 2 Silver , Copper-Antimony ( commercial wire ) Silver .
010 Mean .
t8 0 t2=33 0 v=1059 ) t 8-0 t2=31 2 v=-P055 &gt; 1 054 t , 80 t2=304 v=1052 Lithium , Silver-2 Silver , Copper .
t5-2 t2=32-5 v=1-888 ' tj= 5-2 t2= 31-5 v=1l887 1 884 t , = 5-2 t2= 300 v=1 8801 t1= 5-2 t2=29'0 v=1.881j Calcium , Silver-2 Silver , Copper .
t= 4.3 t2=28-8 v=2-132 ) t , 43 t2=28O0 V=2'129 2-130 tl= 4.3 t2=26i9 v=2.129 Iron , Silver-2 Silver , Copper .
t= 6 , O t2=28-8 v=2.616 ) t= 6-0 t2=26-2 v=2-608 92609 t , = 6-0 t2=29-8 v=2-604J Antimony ( pure wire ) , Argentine-Lead , Argentine .
t1= 9.5 t2=35.1 v=1 234 ) t1= 9.5 t2= 33 9 v=1*236 &gt; 1*236 t1= 9.5 t2=32-9 v=1-237J Copper ( galvanoplastic ) , Copper-Silver , Copper .
t , =26-2 t2= 480 v=1 240 ) t , =26 2 t2=46-5 v=1-246 1'244 t1=26 &gt; 2 t2=45-7 v=1247 ) Strontium , Argentine Silver , Argentine .
t1= 6.0 t2= 36-1 v=1'383 ) t1= 60 t2=34-7 v=1 390 1 387 t1= 6-0 t2=33-1 v=1 389J All the temperatures are given in Celsius 's degrees .
For the first experiment of the series , a thermo-element formed out of commercial bismuth ( pressed wire ) , and silver was compared with a thermo-battery composed of five silver-copper elements .
The silver and copper were the same sorts as those whose thermo-electric constants were taken =0 and 1 .
The temperature ( t , ) of the water in the box containing the cold soldering-points was 7 ?
6 , and that of the warm ( t2 ) , 240 0 .
The ratio of the electromotive force of the two elements ( v ) was found =7'163 , in two other experiments , 7-164 and 7-157 .
The .
mean of these three numbers is 7-162 , and therefore the electromotive force of a bismuth-silver element = 35-81 , that of a silver-copper one , or in other words , the thermo-electric constant of commercial bismuth ( pressed wire ) , = 3581 .
All the other numbers in the Table have a similar meaning , and the metals , whose thermoelectric numbers are calculated out of them , are written in italics .
The connexion was , in all cases where the metal allowed of it , formed by soldering .
Antimony wire , which ( as well as that of tellurium ) is so brittle that it will hardly bear touching , could of course not be directly soldered , but had to be treated in the following manner .
The wire was pressed in a glass tube , in which it was soldered to two thick copper wires with ROSE'S metal ; these , in order to keep them from moving , were made fast in the tube with gypsum .
Lithium , calcium , strontium , and magnesium were experimented with in sealed-up glass tubes filled with rock oil , as fig. 1 shows , the ends of the wire being held at a Fig , 1 .
and b by two steel wire-holders , whose jaws press so strongly together that they flatten out the ends of the wire .
At the other end of the wire-holders platinum wires were soldered on , which passed out of the tube at c and d. The way in which this piece of apparatus was put together is as follows:-a long platinum wire , soldered on to one of the wire-holders , was passed through a glass tube ( drawn out and bent as shown in fig. 2 ) atf , so that the end of it came out at g , whilst the wire-holder remained outside Fig. 2 .
the other end atf .
The glass tube being then laid in a trough filled with rock oil , the wire , previously pressed in a tube filled with the same liquid , was then taken out of the tube in the trough , the one end of it scraped clean and fastened in the wire-holder , and drawn into the tube by means of the long platinum wire , until only the other end of it remained out ; this also having been scraped clean and fastened in the second wire-holder , was drawn into the tube as far as fig. 2 shows .
The tube was now lifted out of the trough in an inclined position , so that it almost remained filled with the oil , and sealed up at g ; aAf it was then drawn out , the oil boiled to drive out the air , and sealed up as shown in fig. 1 ; and lastly , the long end of the glass was melted off at c , and the superfluous platinum wire cut off .
During the experiments the tube was so placed that the one end found itself in the middle of the one , and the other in the middle of the other box ; that filled with warm water was placed higher than the other , in order that the tube might remain in an inclined position , preventing thereby the warming of the other end through currents which would otherwise have arisen in the oil .
The connexion was now formed by soldering on the silver or other wires to the platinum ones .
To ascertain whether this arrangement would give correct results , a silver-palladium element was compared with two composed of silver-copper ; in the one case the palladium wire was fastened in the wire-holders as above described , in the other it was soldered directly to the silver .
The ratio of the electromotive power of the two elements was found , when the wire-holders were used , in four observations , 1 779 , 1 776 , i 767 , 1777 , the mean of which is 1'775 , agreeing very well with 1'780 , the mean of the results obtained when the wires were soldered together .
The above-described boxes could only be used with metals in form of wire , or rods of a considerable length ( at least 8 inches ) ; for experimenting with other metals two smaller boxes were employed , of which fig. 3 is a sketch .
A and B are two copper boxes of Fig. 3 .
about 4 inches in height , 4 in width , and 5 long .
B consists of two halves , a and b , insulated from each other by a piece of india-rubber .
A was filled with warm , B with cold oil .
If two pieces of different metals , c and d , are placed between the two boxes , which are pressed against each other by screws , and if the two wires e and f , soldered to the two halves of B , be connected , a current is formed in the circuit whose electromotive power is equal to that of a thermo-element constructed with the metals c and d , whose extremities have the temperature of the oil in the boxes .
To determine this electromotive power , a circuit was formed with the wires of the multiplier used for the former experiments , the two wires e and f , and the wires of one or more thermoelements made of wires whose soldering-points dipped alternately in the oil of each box .
The two commutators were again brought into the circuit , and answered the same purpose as before .
Where two or more thermo-elements were used , as was generally the case , they were fixed in a piece of wood , as Fig. 4 .
fig. 4 shows , and which needs no explanation .
With the same apparatus the difference in the thermoelectric behaviour of the cleavage planes of bismuth and antimony crystals was studied .
The bismuth crystals had about 8 millims. the sides ; two cleavage planes were made perpendicular to the principal axis , and two parallel to it .
As antimony crystals could not be obl tained large enough for these experiments , pieces were cut out of commercial antimony , as FRANZ did for his research on the same subject* .
These were so experimented with , that in one case the planes perpenaicular to the principal axis , in the other those parallel to it , were pressed between the sides of the boxes .
The position of the crystals , where the current passes in the direction of the principal axis , is called the axial , the other the equatorial .
The results of the observations with the crystals do not agree very well with each other , owing probably to the cleavage planes not being exactly ground at the proper angles , and also to their being placed in a rather oblique position between the boxes .
Table III .
gives the results obtained with the copper boxes .
TABLE III .
7 Copper , Argentine-Argentine , Bismuth ( pure cast ) .
t1=25 5 t2=47 0 v=1 503M t1=25 8 t2=404 v=1.509 1 505 tl=25-5 t2 =38.7 v_=1-504J Silver , Cobalt ( No. 1)-Silver , Argentine .
t =18-7 t2=43 7 v=1 722 ) t =18-7 t2=400 v=1710 1713 t , =19-3 t2=36-8 v=1-708J Silver , Argentine-Silver , Nickel .
t=-205 t2= 56-0 v-=1O040 ) t_-206 t2 =502 v=1-048 1-044 t1=21-0 t2=44X6 v=1-044 Silver , Argentine-Silver , Cobalt ( No. 2 ) .
t -21-2 t2=600 v=1.389 ) t , =21A4 t2=53-0 v=1-409 1-398 t -21-7 t2=47-7 v=1-395J Silver , Argentine-Iridium , Argentine . ?
o Meanm t1=22 5 t2=498 v-1029 ' t , =22-7 t2=46-8 v=1 039 1032 t1=232 t2_453 v=1027J Gas-coke , Argentine-Silver , Argentine .
t =26i7 t2=57 5 v=1O010 ) ti=27 2 t2-56A4 v=1 012 1O011 t1=27 2 t2=51 9 v=1'012J Silver , Argentine-Arsenic , Silver .
t1=23 7 t2=43 3 v=1P368 ) ti=24 Q t2=40 5 v=-1372 &gt; 1369 t , =24 4 t2=36 9 v=1P366J Antimony ( pure cast ) , Silver-Silver , Argentine .
t1=22 3 t2=578 v=1 884 ) t1=22 5 t2=53 5 v=1 882 p1884 t1=23d1 t2-48 2 v-1886J Tellurium , Silver19 Copper , Argentine .
t1=17 0 t2-36 2 v=2 2591 t1=17'5 t2=326 v=2235 ti=18 3 t2=29 7 v=2 215 2 232 t =18.5 t2= 375 v=2 214 t1=18 8 t2=32 1 v=2 235 t1=19K1 t2-306 v=2228J Silver , Bismuth-Antimony-Silver , Argentine .
t1=24 6 t2=58 2 v=5 525 ) ti=24 8 t2=56A4 v=5 575 5 546 t , =25 2 t254 1 v=5 538J Bismuth-Tin , Silver-Silver , Argentine .
t1=26 2 t2=56 5 v=2.609 ) t1=266 t2=5316 v-2617 2'608 t , =27 0 t2=51L0 v=2 598J Antimony-Zinc , Silver-Silver , Argentine .
t , =24-3 t. 515 v=4 329 ) t1=24 6 t2=49-3 v=4-325 4-333 t1=25-0 t2==475 v=4-345 5 Silver , ArgentineSilver , Bismuth ( crystal axial ) .
t , =30 0 t2=45.6 v=1152 ) t1=30 5 t2=44 6 v=1P150 1I154 t1=3I2 t2428 v=1K160j Silver , Argentine-Silver , Bisnuth ( crystal No. 1 , equatorial ) .
ti -31f2 t2 58@1 v=1'7O6M t -32'5 t2=48-7 v-1-700 1-702 t =32-5 t2 48-1 v=1-700J 5 Silver , Argentine-Silver , Bismuth ( crystal No. 2 , axial ) .
t1=300 t2=44-3 v=1 073 ) t , =300 t2=42-7 v=1'070 1 075 t -302 t= 41 -2 v=1.081J 5 Silver , Argentine-Silver , Bismuth ( crystal No. 2 , equatorial ) .
t=28-7 t2=39-5 v=1.492 ) t1=287 t2=39 5 v=1 494 1 499 t29-0 t2=51 0 v=1 500J Silver , Bismuth ( crystal No. 3 , axial)-5 Silver , Argentine .
t -26'2 t2=45-1 v=1-014 ) t -26-3 t2=43-4 v=1 012 1 012 t1=27-5 t2=408 v=1-012J 5 Silver , Argentine-Silver , Bismuth ( crystal No. 3 , equatorial ) .
1t=28 7 t2=400 v=1.446 ) t4=28-7 t2=39-3 v=1-441 &gt; 1-442 t1=28 7 t2=38-7 v=1 438J 5 Silver , Argentine-Silver , Bismuth ( crystal No. 4 , axial ) .
t1= 27-0 t2=502 v=1-660l6 t4=27 4t =46-6 v=1'662J 5 Silver , Argentine-Silver , Bismuth ( crystal No. 4 , equatorial ) .
t1=27-6 t2=50-2 v=1 243 1.240 t1=27 8 t4=46-6 v=1P237J 2 Silver , Argentine-Antimony , No. 1 ( axial ) , Silver .
t , =23 8 t2=34-5 v=1-485 ) t=23 8 t2=38-0 v=1 500 1 493 t , =24O0 t2=35-5 v=1.495j 2 Silver , Argentine-Antimony , No. 1 ( equatorial ) , Silver .
t , =22 4 t2=39 0 v=1-079 t -22-4 t2=36-5 v=1O090 1090 t1=22-8 t2=35-7 v=1-103J 2 Silver , Argentine-Antimony , No. 2 ( axial ) , Silver .
t , =22 6 t2=46 2 v=1 521 ) t -228 t4=_43 4 v=1 517 1-517 t =23 0 4=-44 8 v=1 514J Silver , Argentine-Antimony , No. 2 ( equatorial ) , Silver .
oM Mean .
t , =23-0 t2=42i0 v= 1314 ) t-=23 0 t2=405 v=1L315 p1317 t1 =232 t2 32 2 v=1.321J As tellurium takes such an extraordinary place in the series , it seemed interesting to make a few experiments with a thermo-battery constructed out of tellurium-bismuth ; and one composed of eight of these elements , whose soldering-points had but 600 temperature difference , decomposed a solution of sulphate of copper .
A thermo-battery , made of 100 pairs of tellurium-bismuth , whose extremities have 1000 temperature difference , will give an electromotive power equal to four of DANIELL'S cells .
Such a thermo-battery I am at present constructing .
Selenium and red phosphorus conduct electricity , but so badly , that the galvanometer used in all the former experiments could not here be used , and in its stead an astatic multiplier , with 20,000 folds of wires , was employed .
As the deflections of the needle were very small , in the calculation the angle of deflection may be said to be proportional to the electromotive forces .
Table IV .
contains the results obtained with the astatic multiplier and the copper boxes .
TABLE IV .
2 Silver , Argentine-Red Phosphorus ( only filed , not coppered ) , Silver .
v=108 2 Silver , Argentine-Red Phosphorus ( coppered ) , Silver .
v=1-07 Senen , Bismuth --19 Copper , Argentine .
v=3 57 3610 V=3-62 3~ From the values of the thermoelectric numbers of the different metals given in Table I. , the ratio of the electromotive power of two elements constructed out of any two of those metals may be calculated .
In order , therefore , to ascertain the accuracy of these values , several combinations were made , and their electromotive powers determined .
In the following Tables the ratio of the electromotive powers for the different combinations is given .
TABLE V. Experiments made with the large zinc boxes and the multiplier .
Ratio .
Found .
Calculated .
Silver , Bismuth ( commercial wire)-Silver , Bismuth , Silver , Bismuth ( pure wire).1 0O97 1.088 Silver , Potassium-Silver , Palladium. .
1-543 1-542 Silver , Argentine-Silver , Palladium. .
1-467 1-472 ( Continued .
) Ratio .
Found .
Calculated .
Silver , Palladium-Silver , Sodium. .
1153 1.151 Copper , Potassium-Copper , Sodium ... 2149 2-145 Silver , Quicksilver-Silver , Aluminium .
P 1967 1-967 Silver , Quicksilver-Silver , Magnesium. .
2129 2-143 Silver , Lead-Silver , Tin .
1027 1-029 Silver , Copper ( No. 2)-Silver , Platinum ... 1280 1 275 2 Silver , Gold-Silver , Tin .
1223 1-226 2 Zinc , Silver-Cadmium , Silver ... 1245 1-253 Copper ( precipitated galvanoplastically ) , Argentine-Aluminium , Argentine .
1370 1 385 Antimony ( pure wire ) , Argentine-Copper , Argentine ... .
1228 1-227 Silver , Palladium-Antimony ( commercial wire ) , Silver ... .
1900 1-876 Strontium , Argentine-Aluminium , Argentine .
1838 1 834 Lithium , Silver-Silver , Palladium .
1056 1 058 Iron , Silver-Calcium , Silver .
1220 P1224 TABLE VI .
Experiments with the copper boxes and multiplicator .
Silver , Argentine-Bismuth , alloy of Bismuth , Antimony .
1330 1 278 7 Copper , Argentine-Silver , Bismuth ( pure cast). .
1198 1 189 Silver , Argentine-Argentine , Cobalt ( No. 1). .
1422 1P403 Silver , Argentine-Cobalt ( No. 2 ) , Argentine. .
3497 3'510 Silver , Argentine-Copper , Nickel .
1293 1-303 Silver , Argentine-Iridium , Palladium .
1570 1P542 Silver , Argentine-Gas-coke , Palladium .
1479 1P448 Silver , Argentine-Arsenic , Copper .
1-104 1P086 Antimony ( pure cast ) , Copper-Silver , Argentine .
2068 2-074 Alloy of Bismuth , Tin , alloy of Antimony , Zinc-Silver , Argentine 1P729 1-723 TABLE VII .
Experiments with the copper boxes and the astatic multiplier .
Red Phosphorus ( coppered ) , Copper-2 Silver , Argentine P1040 1P010 Red Phosphorus ( coppered ) , Argentine-2 Silver , Argentine 1-390 1P410 Selen , Antimony-19 Copper , Argentine ... 4370 4-370 Selen , Tellurium.-19 Copper , Argentine ... 2510 2 610 In conclusion , I may here be allowed to express my thanks to Professor KIRCHHOFF for his valuable advice and assistance whilst carrying out this research , and also to Professor BUNSEN for his kindness in having placed his collection of metals at my disposal .

