On the Magnetic Susceptibility Oxygen .
311 ur given in the third column , both the commencement and the raination of the initial change are retarded , the propagation rate iiis nerve at a temperature of G ' C. being slowed to 6 metres per nd .
A comparison of the fourth and fifth columns shows the rdation in the anodic as compared with the acceleration in the odic extrapolar region .
Finally the time relations and the dive E.M.F. of the prolonged effect present in the instances given lie second and fourth columns may be compared with those of I initial change present in all the examples .
It will be seen that change producing the prolonged tail of the photographic record no which differs from that producing the initial spike in the u owing important particulars : it develops slowly , taking from 6 to 0 01 second to culminate , its maximum E.M.F. is only one-\lt ; h of that of the initial change , and it subsides slowly .
It is not , ent in the instances given in columns I , III , and V. i a more extended communication the authors hope to bring / ard other features of the response of nerve , particularly the .acters exhibited by the records of the changes produced by a os of stimuli and of those produced during reflex discharge of the * ral nervous system .
i the Magnetic Susceptibility of Liquid Oxygeu .
" By J. A. Fleming , M.A. , D.Sc .
, F.R.S. , Professor of Electrical Engineering in University College , London , and James Dewar , M.A. , LL. D. , F.R.S. , Fullerian Professor of Chemistry in the Royal Institution , London , \amp ; c. Received May 9 , \#151 ; Read May 12 , 1898 .
i a previous communication * we have described the initial stigations made by us to determine directly the numerical value *ie magnetic permeability of liquid oxygen , and we there indi-d that we hoped before long to present to the Royal Society the its of other experiments made by a different physical method ch we anticipated would enable us to state whether liquid oxygen a constant magnetic susceptibility , or whether , like a ferro-qnetic substance , its magnetic susceptibility alters when subjected ifferent magnetic forces .
Te have recently obtained results which , though limited to a ' ain range of force , we think afford fairly trustworthy values ot magnetic susceptibility of liquid oxygen under magnetising forces ying from 500 to 2500 C.G.S. units , and place them therefore on Lrd .
' Roy .
Soc. Proc. , ' 1896 , vol. 60 , p. 283 , " On the Magnetic Permeability of nd Oxygen and Liquid Air .
" 0L .
LXIli .
^ A Profs .
J. A. Fleming and J. Dewar .
The method used by us in these last experiments depends oitl , well-known fact that if a body , either paramagnetic or diamag ; ( -jc is placed in a magnetic field of variable strength , it is subjected 0# mechanical force tending to displace it in the direction in which } n field varies most rapidly .
If the susceptibility of the body is so small that it doestr sensibly disturb the distribution of the field , the measurement oil .
; mechanical force may be made to furnish a knowledge of the iso-lute value of the magnetic susceptibility .
The necessary conditions are , however , that the volume ( Vjjjtfhe body must be of such small magnitude relatively to the form o h\lt ; field that its magnetisation is not sensibly different from that vick it would obtain if immersed in a uniform field , and also tharhf magnetic susceptibility ( k ) of the substance must be of small absate value .
Under these circumstances , if / is the mechanical bee ; ( reckoned in dynes ) acting on the body , and H is the strengi of the field at its centre , then the force in the direction is giveitbv the equation * f-TcY dx The value h thus determined is a difference value , that is , tis equal to the difference between the susceptibility of the body ad that of the medium in which it is immersed .
Hence if one ni the same body is placed in the same divergent field , but alternalv surrounded by different media , the difference in the apparent susiliabilities of the body in the two cases will give us the differenc e the true susceptibilities of the media .
The experimental method * ployed by us consisted , therefore , in determining the forces acting ?
a small sphere of known susceptibility when suspended first in f and next in liquid oxygen , and hence deducing the difference of 1 susceptibility of liquid oxygen and air , and therefore the abso 8 value of the susceptibility of liquid oxygen , knowing that of air .
The first step was the construction of an electro-magnet capabl 1 producing the required field .
From the above-named conditions^ success it will be seen that since the volume and susceptibility to remeasured are both small , it is essential that the magnetic field si1 not only be large but must vary very rapidly , or else the forces to measured will be small .
An electro-magnet giving the required field was therefore design1 as follows:\#151 ; * ; -aj The excjting bobbin consists of a single coil of double cotte covered copper -wire , Ho. 14 S.W.G. in size .
The coil is 30 cm .
long , 18'5 cm .
in outside diameter , and \#171 ; aperture in the coil 9'5 cm .
in diameter .
The total weight On the Magnetic Susceptibility of Liquid Oxygen .
313 ble cotton-covered copper wire on the bobbin is 71 lbs. , and the ,1 length of wire is roughly about 3300 feet .
The total number of .
)S of wire on the bobbin is 2478 .
The wire is capable of carry-12 or 14 amperes for considerable periods of time without j gerous heating .
The total resistance of the wire is very nearly s ohnis at 20 ' C. his bobbin of wire is enclosed in a cylinder of mild steel the same height as the bobbin , the walls of which are cm .
thick .
This cylinder has movable circular end plates of l 2'5 cm.^in thickness fitting it , and in these plates are circular i\#171 ; s 9 cm .
in diameter .
A soft steel core was also provided , mmm , vacuum vessel ; B , ball weighed ; W , wire windings of magnet ; S , C , steel shell and magnet core .
cm .
in length and 9 cm .
in diameter , and having steel 3tuds '.awed into it so that the core could be held in he position own in the diagram .
The end surface of the inner core was \#166 ; \gt ; cm .
below the upper surface of the wire coil .
The steel used high permeability magnet steel .
314 Profs .
J. A. Fleming and J. Dewar .
On passing a current through the magnet coils a magnetic is created in the space just above the upper end of the iler steel core , which is very divergent , and which is not only shn , hut varies very rapidly in an axial direction .
This magnet was placed underneath a shelf on which stood a w sensitive short-beam chemical balance capable of weighing to ij .
tenth of a milligram .
From one pan depended a fine wire pasng through a hole in the balance case and shelf , and served to susjnd a ball just within the field of the magnet .
If the ball was counterpoised by weights and the magnet fen excited , the ball was subjected to an upward or downward ffee which either decreased or increased its apparent weight , aceortnp as it was diamagnetic or paramagnetic .
If W is the gain or losin weight on exciting the magnet , V the volume of the ball , H ie strength of the field at its centre , and the rate of changot the field in a vertical direction , then the equation 981W = AyYH ~ dx gives us the value of the apparent magnetic susceptibility ( \amp ; i)n air of the body weighed .
In order to apply the method we require to know the value of \lt ; field H at different parts along the vertical axis of the magnet f i also the value of the field for various exciting currents .
A careful preliminary investigation on this point was therefe made .
Constant currents varying from 1 to 12 or 14 amperes we passed through the magnet coils and by means of a small second* .
' coil attached to a calibrated ballistic galvanometer the field streBfi at various points in the vertical central axis of the magnet v measured .
These measurements extended from close contact wi* the magnet core to a point about 10 cm .
above the end surface the recessed polar end of the inner steel core .
The distance in centimetres of any point from the end polar st face of the inner steel core is denoted by x , this measureme being made as exactly as possible along the central vertical axis the magnet .
The strength of the field at this point in C.G.S. uni is denoted by H , and the space rate of change in the x direction 1 dH/ dx .
The following tables give the exciting currents ( A ) in ampor measured by a Weston ammeter , No. 3134 , and the values of x and 1 The value of dH/ dx at any point can be at once determined fro1 the curve of H in terms of x. On the Magnetic Susceptibility of Liquid Oxygen .
315 de I.\#151 ; Magnetisation Curves of Tubular Electromagnet , the Induction Density or Field being measured at a point x centimetres from the Pole Face on the Axial Line .
Exciting urrent by Weston ammeter , No. 3134 .
True current in amperes .
A. Induction density ( H ) at various distances x from the pole .
x = 0 *15 cm .
| x = 2*70 cm .
x = 5 *25 cm .
15 / 15*15 4374 14 14 *141 4200 \#151 ; \#151 ; 13*9 14 040 \#151 ; 3269 \#151 ; 13 13 *127 4054 3194 \#151 ; 12 12 *094 3848 3070 2002 *5 11 11*124 3648 2888 1897 10 10*075 3407 2712 1782 *5 9 9*082 3194 2543 1682 8 8*059 3014 2310 1557 7 7.059 2764 2153 1418 *5 6 6*020 2504 1957 1273 5 5*015 2220 1740 1139 4 3*968 1884 1459 955*5 3 2*994 1481 1161 749 2 1*958 992 767 491 *5 The induction density or field along the axial line was also sasured for two constant excitations corresponding to currents of 094 and 6-020 amperes at various distances along the axial line , d the results are given in Table II .
The results in Tables'I and II were then set out carefully in a t 've , and from these curves an interpolation table constructed , owing the axial field at various points on the axis for different citing currents , and these values are given in Table III .
Having calibrated the magnet , the following objects were pro\ ided be used as bodies to weigh in the field , viz. ( 1 ) A silver ball , ( 2 ) a rather smaller copper ball , ( 3 ) several llow glass balls containing a little mercury , and ( 4 ) a bismuth 11 .
In order to test the method and obtain confidence in the results , \\e ade a number of preliminary measurements on the magnetic suscep-oility of water , solutions of manganese sulphate , and ferrous sulphate known densities .
These experiments were made by weighing the silver ball or hollow ass ball when immersed in these liquids contained in a ta vtr laced in the field cavity of the magnet .
Five weighings were always made ( 1 ) The weight of the ball in air , magnet not excited .
( 2 ) The apparent weight of the ball in air , magnet excited .
Profs .
J. -A .
Fleming and J. Dewar .
( 3 ) The apparent weight of the ball in water , magnet not ex ed ( 4 ) The apparent weight of the ball in the solution used , not excited .
( 5 ) The apparent weight of the ball in the liquid , the mjnet being excited with a known current , and the position otthe ball in the field accurately known .
From these weighings we obtain all the required information , 'or from ( 1 ) and ( 3 ) we obtain the volume of the ball , and from ) , ( 2 ) , and ( 3 ) the magnetic susceptibility of the ball , and from ( 4)mi ( 5 ) the magnetic susceptibility of the liquid .
The sphere used as a testing substance was suspended by a mg fine platinum or gold wire from one pan of the balance .
A sail beaker filled with water or the salts solution under test was pLed* over the pole of this magnet and the sphere suspended in it .
- $ In some cases when using strong exciting currents we four it Table II.\#151 ; Induction Densities ( H ) of Field of Tubular Electr magnet at various distances x along the Axial Line .
Distance from pole face = x cm .
Field = H in C.G-.S .
units , For 6 '020 amperes .
For 12*094 amperes .
0*15 0-58 l'Ol 1'44 1'87 2-30 2- 73 3- 16 3- 59 4- 02 4'45 4-88 5'31 5*74 6 17 6'60 7 03 7'46 7'89 8'32 8-75 9 10 9'61 10'04 10 '47 10 91 11 '33 11-76 1346 1253 1151 -5 1052 965'5 867 770 680 596 509'5 425 356 '5 294'5 231 '4 189*9 142 1 114'2 On the Magnetic Susceptibility of Liquid Oxygen .
ible III._Summary of Results of Field Measurements of Axial Field of Tubular Electromagnet .
irrent by feston imeter , \#187 ; .
3134 .
Distance from pole face on axial line .
0 #15 cm .
2 * 70 cm .
5* 25 cm .
0 0 0 0 I 2 994 f 762 492 3 1476 1148 742 4 1778 1460 952 5 2220 1726 1128 6 2492 1950 1274 7 2760 2158 1422 8 2996 2346 1556 9 3226 2534 1678 9 3226 2534 1678 10 3438 2712 1782 11 3650 2880 1890 12 3840 3030 1982 13 4028 3178 2080 14 4208 3320 2168 14-5 4292 3390 2210 15 4374 3460 2254 Ratios of induction density or field H to field corresponding to 6 *0 or 12 *0 true amperes .
0 '15 cm .
2 *70 cm .
5 '25 cm .
0 0 0 0 -3989 0-3908 0 -3862 0 -5923 0-5887 0-5825 0-7135 0-7487 0 -7473 0-8908 0-8852 0-8854 1-0000 1-oooo 1 -oooo 1 -1075 1-1066 1 -1161 1 -2022 1-2031 1 -2213 1 -2945 1-2995 1 -3171 0-8402 0 -8363 0-8466 0 8954 0-8950 0 -8991 0-9506 0 -9504 0-9536 1-0000 1-0000 1 -oooo 1*0490 1-0488 1-0494 1 0959 1-0957 1-0939 1-1177 1-1188 1-1153 1 -1391 1-1420 1 -1372 1 cessary to pack round the beaker of liquid with crushed ice in ler to keep the temperature of the water or solution from being ised by radiation of heat from the magnet coils .
Any rise of tem-rature in the liquid under test , by causing variation in density , troduces a difficulty in obtaining exact and comparable weighings the ball .
; .
The position of jhe ball in the magnetic field was very carelu y liusted , so that the centre of the ball was as nearly as possibie on te axial line of the magnet .
The distance of the centre of the ba om the end of the magnet core was also measured with as muc ; curacy as possible when the balance was lifted off its supports , an st the balance index was in the zero position .
The distance from the pole face.measured in centimetres is called le distance from the pole in the following tables .
Corresponding to his distance ( x ) we know from the preliminary experiments and the esults given in Tables I , II , and III the strength of the magnet eld ( H ) and its axial rate of variation dH/ dx .
\#166 ; The dimensions and weights of the balls used were as follows : Profs .
J. A. Fleming and J. Dewar .
1 Mean o Ball ' made of Weight in grams .
Diameter in cm .
Density .
Volume in c.c. at 15 ' C. Volume in c.c. at \#151 ; 182 ' C. efficient cubica expansic betwee 0 ' C. an \#151 ; 182 ' ( Silver ... ... . .
132 -010 37 -610 62-220 2-90 2-00 2-27 10 -334 q .qo 12 -775 4-226 6-326 12-654 4-190 6-287 0 -000051 Copper ... ... .
Bismuth 9-836 0'000042 Glass ball con- 0 " 000032 taining mercury , No. 1. .
Glass ball con- 18 -663 2-32 \#151 ; 6-539 6-513 0 -600020 taining mercury , No. 2 . .
3-400 1-40 1-462 1*456 Glass ball con- ft m taining mercury , No. 3 . .
32 -500 2-68 10 -343 10 *340 Glass ball con- ft H taining mercury , No. 4. .
5-95 103 \#151 ; 0*588 i 0*586 1 ft In each case a measurement was made of the magnetic suscej-bility in air of the ball used , by weighing it in a known magnee field and observing the loss or gain in weight .
In this manner the following determinations were made:\#151 ; The Magnetic Susceptibility of the Silver Ball .
Weight of silver ball in air at 15 ' C. = 132-010 grams .
Weight of silver ball in water at 2 ' C. = 119-235 grams .
Yolume of silver ball at 15 ' C. = 12-775 c.c. IV .
0- 0132 1- 83 1-78 The absolute magnetic susceptibility of air is 0 024 x 10~6 .
The mean apparent susceptibility of silver in air is therefore fro\#174 ; Loss in weight in grams of silver ball when weighed in field = W ... ... ... ... ... ... ... ... Exciting currents of magnet in amperes ... ... ... ... . .
Magnetic field in C.G.S. units 38 M pVt* I* * Field variation = dll/ dx Distance of centre of ball from pole in cm ... ... ... ... ... .
Calculated apparent susceptibility = \gt ; t110~* 6 i. ir .
hi .
0-0287 0 *0284 0 0142 12 12 6 3320 2820 1798 426 459 308 1*96 3 18 3*18 1*56 1*08 1*97 On the Magnetic Susceptibility of Liquid Oxygen .
above figures equal to \#151 ; 175 X 10-G , and the absolute value 73xl0"6 .
\#166 ; 1 .
Becquerel gives \#151 ; 174 X 10_G as his observed value for silver .
The Magnetic Susceptibility of the Bismuth height of bismuth ball in air at 15 ' C. = 62720 grams .
Volume of bismuth ball at 15 ' C. = 6'326 c.c. / I. II .
nr .
oss in weight in grams of bismuth ball when weighed in field = w . . .
0-085 0-067 0-030 xeiting current of magnet in am- 14 5 12 6 2383 2135 1355 415 372 236 distance of centre of ball from pole \#151 ; gp 2xi cm .
... . .
... ... . .
... . . . . .
4-86 4-86 4-86 alculated apparent susceptibility = jpin-s -13-4 -13 -3 -14-6 The absolute magnetic susceptibility of air is 0 024x10 G. The 3an apparent susceptibility of bismuth in air is , therefore , from the ove figures equal to -1377 xlO"6 , and the absolute susceptibility = \#151 ; 1375 x 10_G .
( Recent values found by ether observers are -13-4x10 ^ Crane ) , and -13-3 xlO-6 ( L. Lombardi ) .
Hence our own is m cordance with these .
* The Magnetic Susceptibility of the Copper Ball .
Weight of copper ball in air at 15 ' C. = 610 giams .
Volume of copper ball in air at 15 ' C. = 4 226 c.c. i. II .
III .
IV .
Gain in weight in grams of cnrmpr Vmll in fipld . . .
0-009 0-005 0 -0055 0 -0035 Exciting currents of magnet in 6 3 6 3 OOQ amperes 1220 709 518 ill 227 132 193 111 Distance of centre cf ball from 5-46 5-46 8-72 8-72 Calculated apparent susceptibility = FlO-6 .
+ 7-6 + 12-5 + 12-9 + 28 320 Profs .
J. A. Fleming and J. Dewar .
The copper ball proved to be ferromagnetic , owing , no doub t traces of iron .
We constructed a curve by which its susceptibi ' in various fields could be deduced from the above observations ?
j the only reason we employed it was because it seemed desirable ' determine the absolute susceptibility of the liquid oxygen with boe\gt ; ' as far as possible different in susceptibility .
Hence we select !
silver , bismuth , and the above slightly ferromagnetic copper ball* the purpose .
The Magnetic Susceptibility of various Glass Balls partly filled \ou Mercury .
1 " ' S| No. 4 ba No. 1 ball .
No. 2 ball/ No. 3 ball .
( nearly f of mercuryi Gain or loss in weight in grams -0*001 6 1330 231 f This ball proved to*^ I be quite neutral j in the strongest field we could | produce J of ball when weighed in field Exciting current of magnet in amperes Magnetic field ... .
Field variation . .
, Distance of centre of ball from pole -0*0044 12 3180 474 -O'OOtf 12 2418 391 of magnet in cm .
Calculated apparent susceptibil- 4*96 \#151 ; 2*38 4*115 ity = \amp ; ilo\#151 ; 6 , . . .
-0*5 zero -0 *277 -1*23 We have , then , the following data for the glass balls i\#151 ; No. J ball .
No. 2 ball .
No. 3 ball .
I No. 4 ball .
1 Weight in grams ... Volume in c.c. at 18*663 3*400 r 32 *500 5*95 15 ' C 6*539 2*32 1 *462 1*40 10 *3427 2*68 0 *5882 1 *03 1 Diameter in cm . . .
, Absolute magnetic susceptibility in vacuum \#151 ; 0 *48 x IQ-\#174 ; \#151 ; 0 *024 x 10~6 \#151 ; 0 *25 x 10"6 \#151 ; 1 *21 x 10^ As a further check on the method we employed the above deter minations of the susceptibilities of the silver and glass ball , No. ^ to obtain the value of the susceptibility of distilled water .
Th\lt ; measurements were as follows:\#151 ; On the Magnetic Susceptibility of Liquid Oxygen .
321 Magnetic Susceptibility of Water . .
, inf susceptibility of ball = \#151 ; 173x10 6 .
[ , mth silver ball i To]ume o\#163 ; baU = m75 ^ tl , w , ... -vy of susceptibility of ball = \#151 ; 0'25 X 10 6 itb glass ball , No. 3 4^ of ^ = 10.3427 CJS_ * i. I. II .
lain or loss in weight ^f ball when weighed in field and in water ... .
Sxciting currents of magnet in am- -0-0146 gram loss 12 -0*0163 loss 14-5 + 0 -0053 increase 12 2802 2848 3180 456 465 474 )istance of centre of ball from pole / up mno-npf , un p.rn. .
3-22 3-79 2-38 apparent susceptibility of ball in wnfpv =r 6 -0-878 -0-945 + 0-333 The absolute susceptibility of water is , therefore , -(1-73-0-88 ) 10-\#174 ; = -0-85x10-\#174 ; ad \#151 ; ( 1'73\#151 ; 0*94 ) 10~* = \#151 ; 079 x 10~\#174 ; from the experiments with xe silver ball ; and -(0-333 + 0-253)10- ' = 0'59 x lO'6 from the speriments with the glass ball No. 3 .
The mean of these values ives \#151 ; 0-74x10- ' as the absolute susceptibility of water .
The allowing are some of the values obtained for the magnetic suscepti-ility of water by older and by more recent observers .
Observer , Faraday E. Becquerel P. Curie Townsend. . .
Quincke Du Bois ... .
Value of k 10\#151 ; '5 for water . . .
0-72 . .
0-67 . .
0-79 . .
0-77 . .
0-81 . .
0-84 Mean value = 077 Hence our value for water 074= H0 " ' is not far from the mean f the above results . .
, , .
Many other experiments were then made with various solutions f salts of iron and manganese , which satisfied us that we could ?
lace reliance upon the results of this method in measuring ie nagnetic susceptibility of a liquid , and we then proceeded to experiments with liquid oxygen .
322 Profs .
J. A. Fleming and J. Dewar .
The balls were accordingly all weighed in liquid oxygen contain in a vacuum vessel placed over the pole piece of the magnet *j-vacuum vessel contained mercury in its vacuum space and was 0 ' " unusually excellent kind .
In it liquid oxygen could be preset for periods of many hours without a trace of ebullition and difficulty was experienced in making the weighings with accuracy .
These weighings of course served also to determine e density of the liquid oxygen used .
The results are embodied in fe following tables .
In each case the weighings give the apparent susceptibility of \#171 ; lquid oxygen , and these figures have to be corrected by addin'ir subtracting a number representing the absolute susceptibility ofs ball at the liquid oxygen temperature .
Thus in the case of i si \ er ball the figure subtracted is two , as the nearest integer repi-senting the susceptibility of silver at \#151 ; 182 ' C. In the case of i bmmuth ball the figure subtracted is sixteen , in the case of the gk balls it is zero or at most unity , and in the case of the copper b the correction is additive , depending on the value of the field .
Table V.\#151 ; Determinations of the Magnetic Susceptibility of Liqui Oxygen .
p ! .
With the Silver Ball .
yolame of ball = 12*684 c.c. at temperature of liquid oxygen .
Density of liquid oxygen = 1*1376 .
Distance of centre of ball from pole of magnet = 5*37 cm .
Magnetic susceptibility of silver ball = 1*73 x 10-6 .
Exciting current by Weston ammeter , No. 3134* .
14 f5 12 9 6 4 3 Field .
H. Field variation^ d H jdx .
Loss in weight of ball in grams .
, W. Apparent susceptibility , ^lO-"6 .
Absolute susceptibility of liquid oxygen , klO~ 2218 408 3 *7585 322 320 2171 400 3 *6242 323 321 1950 359 2 -8845 319 317 j .1646 302 *5 2 -0365 317 315 | 1242 228 1 *2303 337 335 | 928 170 -4 0 -6855 336 334 \#166 ; ' 723 132 *7 0 *4243 343 341 Mean value of susceptibility of liquid oxygen as determined with the silver ball = 326 x 1CT6 .
On the Magnetic Susceptibility of Liquid Oxygen .
3*23 II .
With the Bismuth Ball .
rolume of ball at temperature of liquid oxygen = 6*287 c.c. )ensity of liquid oxygen = 1*1397 .
)istance of centre of ball from pole = 4*865 cm .
fa'*netic susceptibility of bismuth ball = 13*75 X 10-G at 15 ' C. = 15*9xl0-c at -182 ' C. Loss in weight of ball in Absolute Exciting !
irrent by Weston Field .
Field variation .
Susceptibility , fcjlO -6 .
susceptibility of liquid oxygen .
mmeter .
; grams .
\amp ; 10~6 .
12 2135 373 1 *655 325 309 6 1356 236 0*705 344 328 Mean value of susceptibility of liquid oxygen as determined with t ; bismuth ball = 319 X 10 " ' .
III .
With the Copper Ball .
Volume of ball at temperature of liquid oxygen = 4*190 c.c. Density of liquid oxygen = 1*140 .
Distance of centre of ball from pole = 5*46 cm .
Magnetic susceptibility ( varies from 8 X 10 6 to 15 X 10 ) .
Exciting current by Weston ammeter .
Field .
Field variation .
Loss in weight of ball in grams .
Apparent susceptibility , 10"6 .
Absolute susceptibility of liquid oxygen , 7clO~6 .
6 1220 227 0 *3721 315 323 5 1080 201 0 *2930 316 325 4 9L1 169 '5 0*2096 318 329 3 709 131 *9 0 *1282 321 333 2 470 87 *4 0*0546 311 326 Mean value of susceptibility of liquid oxygen as deteimined ie copper ball = 327 x 10-6 .
321 Profs .
J. A. Fleming and J. Dewar .
IV .
With the Glass Ball No. 1 .
Volume of ball at temperature of liquid oxygen = 6*513 c c Density of liquid oxygen = 1*1391 .
Distance of centre of ball from pole = 4*9(j cm .
Magnetic susceptibility of glass ball = 048 x 10-6 .
Exciting current by Weston ammeter .
Field .
Field variation .
Loss in weight of ball in grams .
Apparent susceptibility , Absolute susceptibilii of liquid oxygen , j h. 10~6 , 1 7 6 3 1480 1330 777 257 231 134*9 0*815 ' 0*671 0*229 323 329 329 322*5 328*5 328 *5 Mean value of susceptibility of liquid oxygen as determined wi ; the glass ball No. 1 = 326 x 10-6 .
V. With the Glass Ball No. 2 .
Volume of ball at temperature of liquid oxygen = 1*456 c.e. Magnetic susceptibility of ball = 0*024 x 10-6 .
Distance of ball from pole .
Exciting current of magnet .
Field .
Field variation .
Loss in weight # of ball in grams .
Apparent suscepti- bility , Absolute susceptibility of liquid oxygen , \amp ; .10-V .
4 *57 14 -5 2500 427*5 0 *4919 310*1 310 ^ 14 *5 2500 427 *5 0 *4916 309 *9 310 m 13 2350 402 0*4333 309 *1 309 12 2240 383 0 *3936 309*1 309 \#166 ; 12 2240 383 0 *3947 310 *0 310 6 1422 236*5 0*1721 344*8 345 6 1422 236*5 0 *1723 345*2 345 7 *04 12 1379 317 0 *2260 348*4 348 12 1379 317 0 *2252 347*1 347 6 865 215 0*0894 323*9 324 W \#151 ; 6 .865 j 215 0*0894 j 623*9 324 , , Mean value of susceptibility of liquid oxygen the glass ball No. 2 = 325 x 10"6 .
as determined with On the Magnetic Susceptibility Oxygen .
325 VI .
With the Glass Ball No. 4 .
'olume of ball at temperature of liquid oxygen = 0-5858 c.c. lagnetic susceptibility of ball = \#151 ; 1*23 x 10~6 .
istanee f ball \gt ; m pole .
Exciting current of magnet .
Field .
Field variation .
Loss in weight of ball in grains .
Apparent suscepti- bility , *xio- ? .
Absolute susceptibility of liquid oxygen , HCT6 .
t-935 15 2392 437 0-1912 306 305 15 2392 437 0-1902 305 304 12 2102 384 0 -1475 306 305 6 1337 234 0-0628 336 335 12 2102 384 0 -1477 306-5 305 6 1337 234 0 -0620 332 331 1-115 12 2418 391 0 -1795 318 317 6 1532 238 0 -0790 361 360 Mean value of susceptibility of liquid oxygen as determined with :Ass ball No. 4 = 320 x 10"6 .
The absolute susceptibility is derived from the apparent suscepti-lity by adding or subtracting the susceptibility of the ball used cording as it is paramagnetic or diamagnetic .
In the case of the copper ball , owing probably to traces of iron , the .ramagnetic susceptibility is rather large and varies with the field , lie proper additive correction was obtained by drawing a curve id setting off the observed values of the copper ball susceptibility ordinates corresponding to the proper field strengths .
The mean value from all the six sets of observations compiising lirty-six determinations with the silver , copper , bismuth , and four !
ass balls is to give a value of 324 x 10~6 as the mean co-efficient of lagnetic susceptibility of liquid oxygen .
From the relation = l+4jrfe , the magnetic permeability can be deduced , and it = 324 x 10-6 we have fi = 1-0041 , as the value of the permeability f liquid oxygen .
The value of / *we gave ( see ' Proc. Hoy .
Soc. , vol. 60 , p. 29- ) s the result of our former experiments by a totally different method .-as \#171 ; = 1-00287 , or nearly , 1-003 .
Hence these two methods agree in giving values of the magnetic lermeability of liquid oxygen differing only by about one part in a housand .
The results of the present work must , however , be taken as giving i much more probable value of the magnetic susceptibility .
On examining the above results it will be seen that there is a Profs .
J. A. Fleming and J. Dewar .
general tendency for the susceptibility with large fields of the c| of 2500 to be less than the susceptibility for fields of the tier of 500 .
The average susceptibility ih fields of from 2500 to 100 is more nearly 310 x 10~6 , and that in fields from 1100 to 500 nearly 330 x lO"6 .
The difference only amounts to about 10 per cent , of the lo^ value , and it cannot be said that the observations are all of exailv equal weight .
The general result is to show that between the limits of H = , \#171 ; ) and H = 2500 the average magnetic susceptibility of liquid oxy.n has a value which does not differ much from 324 x 10~8 , but wit small but decided tendency to decrease in strong fields .
The determination of the variation of susceptibility in mih weaker fields is left undecided by these experiments , but by te employment of a torsion balance we hope to be able before long \#187 ; give the latios of the susceptibility in various fields much weal* than those employed in the foregoing experiments .
In connection with the determination of the absolute magnet ' susceptibility of liquid oxygen , our attention has been much direct to the important matter of the determination of magnetic suscep* bilities of substances in general at very low temperatures .
Haying regard to the great loss in magnetic suseeptibili experienced by the ferromagnetic metals in heating beyond a certa .
temperature , it has been frequently suggested that bodies of sma susceptibility might become strongly magnetic if cooled to sufficiently low temperature .
Faraday made many experiments o this question , using solid carbonic acid as a refrigerating agent , bu was not able to arrive at any conclusions .
A difficulty which presents itself in the use of liquid oxygen a a refrigerating agent for this purpose is the strongly magnetic quality of the liquid itself .
All bodies except iron , nickel , an\lt ; cobalt , and the strongly ferromagnetic bodies , become apparently diamagnetic when placed in liquid oxygen and in a non-uniform magnetic field.* Moreover , for obvious reasons it is easy to weigh a diamagnetic ot apparently diamagnetic body in a non-uniform field because the forces restoring the disturbed body to its original position increase with * An interesting experiment was made with a ball of ebonite which illustrates this fact .
Ebonite is slightly magnetic in air , owing no doubt to iron impurity .
Ebonite is denser than liquid oxygen .
Accordingly , a small ball of ebonite -dropped into the liquid oxygen contained in the vacuum vessel of the pole of the magnet sunk to the bottom of the vessel .
On exciting the magnet the ebonite became apparently strongly diamagnetic and was repelled by the pole .
It floated up in the liquid until it reached a level at which the diamagnetic repulsion just balanced the apparent weight of the ball in the liquid oxygen .
Instructive lecture experiments can be made in this manner .
On the Magnetic Susceptibility of Liquid Oxygen .
327 ii displacement .
If , however , an attempt is made to determine the ce acting upon a paramagnetic body in a non-uniform field by a !
ance , the body weighed is in unstable equilibrium .
We have only recently overcome these difficulties .
The method v have adopted for cooling the body under test , is to suspend it ely near the bottom of a test-tube , which is placed in a vacuum , sel , the interspace between the two being filled with liquid air .
this way the body is cooled by radiation to the temperature of lid air , and yet it is suspended in , and surrounded by , gaseous air , magnetic susceptibility of which is exceedingly small compared :h that of liquid oxygen or liquid air .
3y limiting the vibration of the balance within small limits by t ) stops , or by gradually varying the field of the magnet with a .bon rheostat , until the field is just able to move the object from a ndard position against the fixed restraining force supplied by a istant counter-balancing weight , we have been able to effect the easurements of the apparent weight of the tested object at a given ; tance from the pole , and in a known field , even though the ailibrium is not stable .
In this way we have made preliminary .periments on the variation in the diamagnetic susceptibility of muth , and of the paramagnetic susceptibility of manganous sulphate i the solid condition .
We made a preliminary experiment by weighing in and out of t3 magnetic field a small closed glass bulb , exhausted of its air th when in ordinary air at the normal pressure and temperature , d then suspended on the dense gaseous air in the inner test-tube , rich is at a temperature of \#151 ; 182 ' C. lying at the bottom of the ner test-tube , placed as above described in a vacuum vessel .
We md the magnetic susceptibility of the den\amp ; e air at \#151 ; 182 ' C. to + 028 xlO-6 , in other words about 10 times the susceptibility air at the normal temperature and pressure .
This number 0'28 quite insignificant compared with numbers of the order of 100 or 0 .
Hence an object suspended in the above described manner , can reduced to the temperature of liquid air without changing the sceptibility of the surrounding medium by an amount which is at l comparable either with that of liquid oxygen , or with the value the susceptibility of bismuth , or of most paramagnetic bodies ch as the salts of iron , nickel , cobalt , manganese , or of palladium , any of the strongly paramagnetic bodies .
In this manner we have made a determination of the change in iramagnetic susceptibility of crystallised manganous sulphate in ie form of powder .
We find that the susceptibility of the salt at 25 ' C. is to that o -182 ' C. in the ratio of 105 to 349 or 1 to 332 .
These centigrade temperatures , 25 ' C. and -182'C , correspond to VOL. LX1H , 2 B On the Magnetic Susceptibility of Liquid Oxyg en .
298 and 91 on the absolute thermometric scale , and it is obv therefore that the paramagnetic susceptibility of solid mangaiug !
sulphate is increased by cooling so that its value varies inversely the absolute temperature , and is increased threefold by cooling to \e temperature of liquid oxygen .
It has been shown by P. Curie , that the above law holds good )r j oxygen , palladium and other paramagnetic bodies at high tempa-tures ; and E. Wiedemann ( Pogg .
Ann. , vol. 126 , p. I , 1885 ) ^ Plessner ( Wied .
Ann. , vol. 39 , p. 336 , 1890 ) have shown that )r limited ranges of temperature , 16 ' to 60 ' , it is true for paramagnic j salts in the liquid and solid condition .
Our experiments show that this law may be valid over very vie limits , and to very low temperatures .
We hope to fully examine tis matter shortly , and to make a full examination of paramagn susceptibility at very low temperatures .
We have also made measurements of the magnetic susceptibiiy of bismuth at the liquid oxygen temperature .
From the mean of three experiments , we find that the dianr-netic susceptibility of bismuth is increased by cooling , and thait I has the following values:\#151 ; Diamagnetic susceptibility of bismuth at 15 ' C. == 137 , \#187 ; " " -182 ' C. = 15-9 , thus showing an increase of 16 per cent , on the lower value .
The diamagnetic susceptibility of bismuth is therefore not increa i inversely as the absolute temperature on cooling .
The above considerations suggest that the very large paramagnes susceptibility of liquid oxygen , which is five or six times great ' than that of a saturated solution of ferric chloride , may in parts due to its low temperature .
Some experiments we have made on the susceptibility of dm* ganous sulphate indicate that paramagnetic susceptibility van directly as the density .
The density of liquid oxygen is 806 times that of gaseous oxyg at 0 ' C. and 7b0 mm./ and its absolute temperature is just ox third .
Hence if the law that susceptibility varies as the density w inversely as the absolute temperature holds good down to the te : perature of \#151 ; 182 ' 0 .
, and over the physical change of state , 1 should expect the susceptibility of the liquid oxygen to be 24 .
( = 806 x 3 ) times that of the gas .
The magnetic susceptibility of gaseous oxygen as determined 1 Faraday , E. Becquerel , and others is 0*13 xl0\#151 ; \#174 ; .
How 2418x0 ' ) X10-\#174 ; = 314 x 10-6 .
Aluminium as an Electrode in .
* 329 he mean value we have found for the susceptibility of liquid x/ en is 324x10"\#174 ; , and many of our values for it are exactly 3ix 10"\#174 ; .
seems therefore possible that for paramagnetic bodies over wide -rts of density and temperature we may find that the magnetic iaeptibility varies directly as the density and inversely as the dilute temperature .
'e desire to add that our thanks are due to Mr. J. E. Petavel ..rv Mr. J. T. Morris , for their assistance in carefully carrying out hBedious work of the ballistic observations , necessary to determine h field of the electro-magnet we have used .
\#166 ; uminium as an Electrode in Cells for Direct and Alternate Currents .
" By E. Wilson .
Communicated by Dr. J. Hop-kinSON , F.R.S. Received May 11 , \#151 ; Read May 26 , 1898 .
his paper deals with the apparent great resistance which alumina offers to the passage of an electric current when used as an l ie in cells containing , for instance , such an electrolyte as alum in sir .
The following are references to papers which deal in whole .a part with this or other properties of aluminium when employed ibu electrode in electric cells .
Wheatstone .
* Roy .
Soc. Proc. ' Read April 26,1855 .
This is the Jhest paper I have found dealing with the metal aluminium in caic cells , but Wheatstone does not appear to have noticed the uirent great resistance mentioned above .
\lt ; eeren .
* Mittheil .
des Gewerbevereins fur vHannover , ' Jahrg .
*'\gt ; , p. 342 .
Reference is made in this paper to Wheatstone 's -seriments .
tiff .
' Liebig 's Annalen , ' 1857 , vol. 102 , p. 269 .
The author of paper points out that nine Bunsen elements were not able to \#166 ; i a current through a cell having aluminium as an electrode .
13 is the first mention of this property I can find .
hicretet .
*Comptes Rendus , ' 1875 , vol. 80 , p. 280 ; also ' Journ. ' . !
tPhys .
, ' 1875 , vol. 4 , p. 84 .
Observed great resistance in dilute i^huric acid due to aluminium plate .
eetz .
* Wied .
Ann. , ' 1877 , vol. 2 , p. 94 .
Supposes oxygen to be i-1 cause of this apparent high resistance .
Winhelmann .
1 Wied Ann. , ' 1883 , vol. 20 , p. 91 .
Wright and C. Thompson .
'Phil .
Mag. , ' 1885 , Part 9 , Series 5 , 7\lt ; 19 , pp. 27 , 116 , 203 .
Call attention to the non-compliance of AWnium with thermochemical data .
Reference is made to the , v'k of Julius Thomsen .
2 b 2

