437 On the Influence of the Nature of the Cathode on the Length of the Crookes Dark Space .
By F. W. Aston , B.A. , B.Sc. , A.I.C. , Trinity College , Cambridge .
( Communicated by Sir J. J. Thomson , O.M. , F.R.S. Received August 23 , 1912 .
) In the work previously communicated by the author upon the subject of the cathode dark space only one metal , aluminium , was employed in the construction of the cathodes used .
It therefore seemed desirable that cathodes of various metals , etc. , should be experimented with in order to find out to what extent the relative values of length of dark space , pressure , current density , and voltage were affected by the material of the cathode .
The cathode fall which is intimately connected with the ratio between current and voltage is certainly widely different for different metals* so that it was thought very probable that the length of the dark space might also be affected .
Preliminary experiments fully justified this expectation , and an apparatus was therefore designed for a systematic enquiry using cathodes of many different materials .
At the outset one was faced with the difficulty that , for reasons given in a previous communication , in order to get absolute values of the quantities measured the cathode should be large and of the " guard-ring " type .
As this would put several interesting substances out of court and involve much labour and complication it was decided to use moderate-sized cathodes of the simplest form and treat the values obtained as relative only .
The method adopted of constructing the apparatus so that cathodes could be changed with the least difficulty and risk of accident is indicated in fig. 1 .
B Fig. 1 .
The discharge tube itself was a thin glass shade , of the kind used to protect statuettes , with a hole blown in the end to admit the rod carrying the anode A. Instead of closing the other end with a thick glass plate , which , even with the most careful heating , would be very liable to crack , a thin glass * J. J. Thomson , ' Conduction of Electricity through Gases , ' ed. 2 , pp. 538 and 560 .
Mr. F. W. Aston .
Influence of Nature of [ Aug. 23 , bulb B was used which could be heated up in a naked Bunsen flame with impunity .
The joint was made with sealing-wax and though it was taken apart and remade over 20 times during the research it never showed any tendency to leak .
Through the flask passed the rod R , at the end of this could be screwed the cathode 0 , the face of which consisted of the material under observation .
Both electrodes were circular discs 10 cm .
in diameter , their distance apart about 11 cm .
For simplicity in calculating the current density the whole surface area of 78'5 sq .
cm .
was used .
This involves an edge error , which , however , is probably much the same for each metal used .
The anode was of aluminium and was not changed .
The gases used were oxygen and hydrogen , chosen , the former on account of the abnormal sharpness of the edge of its dark space , the latter for contrast and because it was one which exhibited the primary cathode dark space.* The measuring instruments employed and the procedure of the experiment were the same as described in the foregoing paper on cylindrical cathodes .
As in all work of this kind it was necessary to wash the cathode and the * tube thoroughly with oxygen , while the current was kept running until the effect of a heavy current upon the colour of the discharge was insignificant .
This is a very delicate test , as the characteristic colour of the discharge in pure oxygen is extremely sensitive to impurities .
The duration of such washing varied enormously with the different cathodes used .
Experimental Results .
The behaviour of the cathodes of the different materials used will now be considered in detail , beginning with the metals in the order of their atomic weights .
As before , the length of the dark space D is expressed in centimetres , P the pressure in hundredths of a millimetre of mercury , V the potential difference between the electrodes in volts , and C the current in fifths of a milliampere .
Measurements were only made under restrictions , viz. : that the current must be considerably greater than that necessary to cover the whole cathode with glow , and no positive column should be present .
Under these conditions all the metals satisfied very nearly the two empirical equations obtained with aluminium cathodes , f so that by plotting I ) against 1/ ^/ 0 and V against fC straight lines were obtained from which approximate values of the constants could be calculated , the equations being D A P Vc .
' v = E+^y^ .
* F. W. Aston , 'Roy .
Soc. Proc. , ' 1907 , A , vol. 80 , p. 45 .
t F. W. Aston , ' Roy .
Soc. Proc. , ' 1907 , A , vol. 79 , p. 80 .
1912 .
] Cathode on Length of Crookes Dark Space .
439 C is here the current density , the unit adopted , as before , being one-tenth of a milliampere per square centimetre of cathode surface .
Magnesium.\#151 ; Cathode a plate of the metal about 1*5 mm. thick made by Messrs. Johnson and Matthey .
The surface was not quite flat , so that the measurements of the length of the dark space are not so accurate as desirable .
When first used it disengaged very large quantities of gas , which appeared from spectroscopic observation to be fairly pure hydrogen , for .
which reason it had to be run for a long time before measurements could be made .
No sputtering was observed with this metal .
The surface sustaining the discharge glowed green apparently by impact of the positive rays , this glow became quite brilliant at high voltages and showed the green magnesium line .
This very interesting effect showed a tendency to weaken after prolonged running with hydrogen , so may be due to a surface layer of oxide .
In connection with this phenomenon the author has often observed that cathodes of aluminium , iron , or platinum which have not been specially cleaned give a brilliant yellow fluorescence when first used .
This fluorescence gives the sodium line and disappears rapidly with continued running ; it is doubtless due to the casual presence of salts of that metal on the cathode surface .
It will be seen from the following figures that magnesium at a given pressure and current density exhibits the smallest dark space and requires the lowest potential of all the metals as yet experimented upon .
Magnesium\#151 ; Oxygen .
p 13 3 10 -o 8-1 6-9 6-0 c. V. D. y. D. V- D. y. D. !
y. D. 10 358 1 -155 373 1 -285 392 1 -420 413 1 -555 437 1 -690 16 363 1 -ooo 390 1 -135 415 1 -260 440 1 -390 470 1 -540 25 377 0-870 408 1 -ooo 440 1 -125 471 1 -250 508 1 -400 44,1 .
398 0-730 442 0-860 487 1 -ooo 527 1 130 575 1 -265 Magnesium\#151 ; Hydrogen .
| P ' ' | 27 -6 25 2 19 -8 17 -0 14-5 C. V. D. !
V. I ) .
y. D. V. D. V. D. 16 I 302 1 *460 310 1 -560 j 353 1-815 383 1-955 432 2-190 25 325 1 *380 341 l -440 i 392 1-660 430 1 -840 490 2-075 44i 370 1 *235 397 1 -325 460 1 -545 510 1 -730 583 1 945 64 410 1 *155 442 1 -255 | 515 1 -485 570 1 -670 658 1 -880 440 Mr. F. W. Aston .
Influence of Nature of [ Aug. 23 , Aluminium.\#151 ; Cathode a plate 3 mm. thick , turned flat in the lathe .
As it is light , easy to work , and carries more current with less sputtering than any of the other metals easily available , its technical use as material for cathodes is almost universal .
For purposes of research in which pure gases are required these advantages are considerably discounted by its unfortunate property of giving off very large quantities of gas when in use .
On this occasion the evolution of this gas , which appears to consist of .carbon compounds , did not wholly cease even after many days ' running .
During such long running its surface appears to become fatigued , but when once thoroughly run with oxygen it gives very consistent results so long as the vapour of mercury is carefully excluded from the apparatus .
Aluminium\#151 ; Oxygen .
p 16 -5 11 *2 9-9 7-8 7 -2 0 .
y. D. y. D. y. D. y. D. y. ] ) .
10 362 1 -200 388 1 -370 400 1 -465 430 1 -625 448 1 -760 16 372 1 -030 405 1 -185 42 L 1 -310 460 1 -450 485 1 -550 25 385 0-910 428 1 -050 456 1-150 496 1 -300 529 1 -430 44| 1 411 0-750 467 0-910 505 1 -020 555 1 -135 600 1 -255 Aluminium\#151 ; Hydrogen .
P 33 -3 26 -0 22 -0 19-3 15 -7 C. j y. D. y. P. y. D. V. D. y. D. 10 367 2-080 430 2-450 16 313 1-375 340 1 -605 360 1 -780 417 1 -940 495 2 -300 25 333 1 -240 380 1-465 411 1 -640 480 1-830 574 2-150 44 ^ 380 1 -105 450 1 -340 490 1 -540 578 1 -730 700 2-090 64 421 1 -045 508 1 -275 560 1 -500 \#151 ; \#151 ; \#151 ; Iron.\#151 ; Cathode a plate of soft iron of the quality used for magnet laminations .
The surface was well cleaned with emery paper ; it gave off very little gas , but sputtered rather badly in hydrogen .
1912 .
] Cathode on Length of Crookes Dark Space .
Iron\#151 ; Oxygen .
p 15 -9 13 .o 11 -o 9-6 0 .
y. D. y. I\gt ; .
V. D. y. I ) .
10 432 1 -270 440 1 -380 456 1 -500 480 1 640 1G 450 1 -115 462 1-230 487 1 -350 517 1 -470 25 46G 1 -ooo 492 1 115 522 1 -230 563 1 -355 44# 500 0-8S0 542 1 000 589 1 -no 640 1 -240 Iron \#151 ; Hydrogen .
P 40-3 32 -6 28 -7 23 -8 C. y. D. y. D. 1 y. D. Y. 1 D. 16 427 1 -625 438 1 -780 468 1 -880 507 2-055 25 451 ; 1 -475 483 1 -635 520 1-750 569 1 -915 44# 490 1 -320 560 1 -485 608 1 -610 676 1 -765 64 j 530 1 -250 620 1 -420 680 1 -540 760 1 -710 Copper.\#151 ; Cathode a plate of the commercial metal about 8 mm. thick , the surface being cleaned with emery paper just before it was put in the tube .
It gave off very little gas , and except for sputtering behaved very well .
Contrary to expectations the surface did not become discoloured by the discharge in oxygen .
Copper\#151 ; Oxygen .
P 13-4 10 -5 9-2 : !
7 -8 C. v ' i D. y. D. y. D. y. D. 10 440 1 -435 477 1 -610 505 1 -740 542 1 -945 16 470 1 -285 514 1 -460 548 1 -580 595 1 -760 25 503 1 -165 559 l -340 598 1 -470 657 1 -625 44# 561 1 -045 632 1 -210 683 1 -330 753 1 -490 Copper\#151 ; Hydrogen .
P 44 -9 i 37 -0 30-8 26 *7 C. !
T. !
1 1 D. V. D. y. D. y. D. !
16 477 1 -835 527 2 -020 596 2 -230 660 2 -435 !
25 517 1 -670 590 1 -860 672 2-105 745 2 -325 1 44# 590 1 -500 690 1-730 800 1 -940 894 2 -175 64 645 1 -420 763 1 -640 897 1 -860 980 2 -100 Mr. F. W. Aston .
Influence of Nature of [ Aug. 23 , Zinc.\#151 ; Cathode a plate of the commercial metal about 1 mm. thick .
This metal behaved remarkably well as a cathode , giving off very little gas , so that the time of running was comparatively short before constant results were obtained .
At the close of the set of readings when taken out it had a faint greyish bloom on the part carrying the current , but no sputtering effect was noticeable on the glass walls of the tube .
Zinc\#151 ; Oxygen .
p 13-1 11 -1 96 7-8 6-3 1 0 .
V. D. Y. D. y. D. V. D. V. D. 10 417 1 -380 440 1 -550 459 1 -630 495 1 -820 550 2-130 16 438 1 -245 470 1 *350 490 l -440 538 1 -640 605 1 -930 25 470 1 -085 504 1 -210 530 1 -315 590 1 -505 670 1 -765 44* 517 0-965 562 1 -070 595 1 -170 668 1 -380 773 1 -620 Zinc\#151 ; Hydrogen .
P 42 -4 39 *2 31 -5 27 -6 C. y. D. V. D. V. D. y. D. 16 378 1-645 408 1 -740 465 2 -010 503 2 -200 25 411 1 -485 450 1 -630 528 1 -890 580 2 -080 44* 478 1 -375 528 1 -500 627 1 -765 697 1 -970 64 528 1 -300 590 1 -430 700 1 -700 783 1 -880 Silver.\#151 ; Cathode thick bright foil stretched upon the copper cathode already described .
When used with oxygen the surface very soon took a brownish tinge , which it lost almost completely in hydrogen .
Silver gave in oxygen larger values of D than any other metal , but sputtered very badly , in hydrogen so badly that very few reliable readings could be obtained .
Silver\#151 ; Oxygen .
P 19 -5 15 -0 12-8 11 -o 9-6 1 C. y. D. y. i d. !
V. D. y. D. y. D. 10 452 1 -330 471 1 -400 480 1 -600 508 1-730 535 1 -870 16 417 1 -185 493 1 -315 512 1-435 548 1 -575 583 1 -715 25 487 1-075 517 1 -200 550 1 -305 596 1 -450 640 1-575 44* 522 0-965 572 1 -090 622 1 -200 685 1 -340 740 1 -470 1912 .
] Cathode on Length of Crookes Dark Space .
Silver\#151 ; Hydrogen .
p 58 -5 48-1 40-0 c. V. D. V. D. V. D. 16 460 1 -600 510 1 -800 560 1 -940 25 490 1-460 552 1 645 620 1 -815 44 !
530 1 -315 630 1 -500 722 1 -670 64 597 1 -220 690 1 -415 \#151 ; \#151 ; Tin.\#151 ; Cathode thick bright foil , used in the same way as the silver .
Sputtered very much less than that metal and did not become discoloured .
It gave off very little gas , and on the whole was the most satisfactory of the heavy metals .
Tin\#151 ; Oxygen .
P 19-0 13 -4 9-8 8-2 C. Y. D. y. D. Y. D. Y. 1 D. 10 436 1 " 220 450 1 -395 483 1 -600 515 1 -755 16 448 1 -070 472 1 -220 520 1 -430 557 1 -610 25 462 0-945 500 1 -100 560 1 -300 605 1 *465 44 !
489 0-810 548 0-970 627 1 -180 687 1 -330 Tin\#151 ; Hydrogen .
P 47-5 39-4 34-4 30-6 C. V. D. V. D. Y. D. Y. !
D- 16 ( 432 ) 1 -520 450 1 -745 487 1 -905 520 2-045 25 ( 465 ) 1-395 487 1 -765 548 1 -765 590 1 -920 44 !
( 502 ) 1 -240 565 1 -450 650 1 -620 705 1 -780 64 540 1 -190 625 1 -355 725 1-545 Note.\#151 ; Readings in brackets are unreliable owing to the presence of a positive column .
Platinum.\#151 ; Cathode thick foil cemented upon the zinc plate already used .
The metal gave off a considerable amount of gas , and sputtered very badly .
Mr. F. W. Aston .
Influence of Nature of [ Aug. 23 , Platinum\#151 ; Oxygen .
p 16-9 12 1 10 -o 9 -4 7 -9 c. Y. D. Y. D. Y. D. Y. D. Y. D. 10 435 1-315 462 1 -480 481 1 -620 508 1 -765 547 1-950 16 449 1 -170 490 1 -335 520 1 -455 552 1 -600 600 1-770 25 471 1 -055 522 1 -205 568 1 -340 608 1-470 665 1 -650 44 j 510 0-915 582 1 -090 648 1 -220 698 1 -350 773 1-515 : .
Platinum\#151 ; Hydrogen .
P 45 -4 37 -9 31 -8 27 -5 c. Y. D. | Y. D. Y. D. V. D. 16 422 1 -715 434 1-855 514 2-130 588 2-315 25 456 1 -565 495 1 -735 583 1 -990 683 2-200 44i 510 1-430 580 1 -595 697 1 -840 808 2 -045 64 562 1 -355 640 1-535 780 1 -750 910 1 -970 Lead.\#151 ; Cathode a thick plate well scoured with emery paper just before use .
The oxygen discharge turned the surface a rich brown , which was reduced to a greyish appearance in hydrogen .
This metal was the worst of all for sputtering , measurements in hydrogen being almost unattainable , and not to be relied on .
Lead\#151 ; Oxygen .
P 15 -3 11 -9 9-3 7-8 0 .
Y. D. V. D. Y. D. Y. D. 10 452 1 -380 467 1 -550 516 1 -730 566 1 -940 16 474 1 -230 505 1 -365 562 1 -570 620 1 -800 25 500 1 -100 548 1 -235 618 1 -440 684 1 -630 44\#163 ; 555 0-955 619 1 -120 708 1 -300 785 1 -480 | Lead\#151 ; Hydrogen .
P 49 -1 39 -7 35 -9 C. Y. D. Y. D. Y. D ' 16 ( 500 ) 1 -880 564 2 -130 628 2 -320 !
25 ( 570 ) 1 -700 656 2-000 708 2-125 44A ( 648 ) 1 " 525 770 1-830 835 1 -985 '64 710 1 -420 850 1 -740 \#151 ; 1 1912 .
] Cathode on Length of Crookes Dark Space .
Experiments on Carbon .
It was thought on account of its extreme position in the Volta series that this element might yield remarkable results .
A carbon surface was first prepared by coating a lead cathode with ordinary graphite stove-polish .
This did not give satisfactory results , so a cathode was cut out of an old gas-carbon battery plate about 5 mm. thick .
This was well washed , dried , and ignited in a muffle , and finally the surface was ground flat with emery .
After long continued running and pumping the gas which it gave off was apparently exhausted , but it proved quite impossible to get reliable readings in oxygen , owing no doubt to the formation of oxides of carbon .
A few single readings , however , show definitely that its dark space length is not greater than that of silver .
In hydrogen some successful observations were made with results given below .
Carbon\#151 ; H y d rogen .
p 52 1 45 -0 34 -2 28-5 c. Y. D. Y. D. Y. D. V. D. 16 ( 422 ) 1 -420 443 1 -520 480 1 -780 537 2-030 25 ( 430 ) 1 *260 458 1 -400 539 1 -640 615 1 -900 44J ( 480 ) 1 -120 510 1 -265 644 1 *550 747 1 -790 64 ( 515 ) 1 *050 555 1 -170 725 1 -500 840 1 -745 Experiments with Liquids .
By means of another and smaller discharge tube some rough measurements were made with liquid electrodes .
The tube was in this case vertical ; the surface of the liquid itself formed the cathode , the anode being an aluminium plate .
Mercury.\#151 ; This metal acted quite well as a cathode when perfectly pure and superficially clean .
It showed no outstanding peculiarities which would warrant the trouble of making a special apparatus for the determination of its constants .
These appear much the same as those of zinc .
Sulphuric Acid.\#151 ; When the mercury was replaced by concentrated sulphuric acid it gave off such volumes of vapour , when used as a cathode , that it was impossible to make measurements at constant pressure .
When the surface was used as an anode , although the volume of gas liberated by electrolysis would be the same , this did not occur ( a very interesting point , since mercury did not behave in this way ) , so that it was possible , by the use of the aluminium disc as cathode , to get the aluminium value Mr. F. W. Aston .
Influence of Nature [ Aug. 23 , for the dark space at the prevailing conditions , and then by instantaneously reversing the current obtain a rough comparative value for the sulphuric acid .
This was found to be about the same as the aluminium one .
Experiments on the Surface Configuration .
In order to ascertain how far roughness , as apart from cleanness , of the cathode surface affected the results , a plate of zinc was taken and the entire surface cut in parallel grooves in a shaping machine .
The grooves were contiguous and about 0'5 mm. deep , so that the actual surface area of the cathode was increased some 50 per cent. A series of readings were taken , when it was found that , so long as D was measured from a point about half way down the grooves , the values of all the constants were practically identical with those for the flat plate .
This definitely shows , first , that mere roughness , even of an exaggerated kind , does not affect the discharge , and secondly , that the current factor B depends on the mean normal current density so long as the irregularities of surface are small compared to D. { Compare with results in previous paper on curved cathodes .
) Experiments with Perforated Cathodes .
The subject of cathodes with discontinuous surface is a very complex one , and one of peculiar interest , inasmuch as the whole of the work on positive rays has been done with their aid .
On this occasion two cases only were considered , which are rather to be regarded as tentative experiments in a new field than properly part of the work now described .
Commercial perforated zinc was used in these experiments .
The perforations were about 1*2 mm. in diameter , and so numerous that they took up approximately half the surface .
In the first case , 3 mm. in front of the zinc cathode already used , a disc of the same size was mounted on thin metallic supports ; if it was put further away , the* discharge struck back through the perforations .
With this arrangement it was found that , notwithstanding the enormous reduction in the surface of metal immediately exposed , the current actually passed more easily than with the plane zinc cathode , streamers of light , doubtless due to positive rays pouring past the perforated plate upon the other , could be seen behind the holes , and the light in front of the cathode was certainly more intense in front of the perforations than elsewhere .
As soon as a series of readings had been taken , a plate of mica was introduced at the back of the perforated plate to prevent the streams of positive rays reaching the plate behind .
The cathode now carried considerably 1912 .
] Cathode on Length of Crookes Dark Space .
447 less current than the plane one , the dark space being correspondingly larger .
These perforated cathodes were the only ones that showed any marked tendency to diverge from the normal behaviour expressed by the two empirical equations already quoted .
The values given to their constants in the tables are only approximate .
Primary Dark Space*\#151 ; All the cathodes used in hydrogen exhibited a primary dark space of approximately the same dimensions with that gas .
Behaviour of the Anode.\#151 ; Towards the end of this research , the anode , which had been used continuously the whole time , gave evidence of a new and interesting phenomenon .
This was the anode dark space described in a letter to ' Nature .
' !
It was thought that the presence of this surface-fatigue effect might interfere with the accuracy of the voltage readings , so the cathode under observation at the time was removed , and the original aluminium one replaced .
Although this did not affect the appearance of the anode dark space , the original values of the voltage , etc. , were given again within the limits of experimental error , a very satisfactory check on any changes in the apparatus which might have occurred during the time over which the experiments extended .
Table of Constants .
Oxygen .
1 A. 1 B. E. F. F/ A. Di- V|* Vi/ D l Mg 4-8 0-41 310 13 -5 283 0-890 434 486 A1 5-7 0-43 310 17 -5 302 1 -035 492 483 Fe 8-5 0-38 337 26 -0 306 1 -230 597 484 \lt ; Ju 8 *9 0-40 340 28 -5 318 1 -290 630 487 Zn 7-3 0-43 335 23 -5 322 1 -170 570 487 Ag 10 -9 0-36 350 33 -0 303 1 -470 697 475 Sn 7-9 0*40 363 24 -0 304 1 -190 603 508 Pt 8-8 0-40 335 30*0 340 1 -240 627 505 Pb 8-7 0-41 340 31 -5 362 1 -280 650 507 Z ' 6-6 0-43 350 ?
24 -0 \#151 ; 1 -128 566 \#151 ; Z " 8-3 0-43 350 33 -0 \#151 ; 1 -290 | 678 \#151 ; ( Z ' , perforated zinc .
Z " , perforated zinc backed with mica .
) * F. W. Aston , ' Roy .
So Proc. , ' 1907 , A , vol. 80 , p. 45 .
t ' Nature , ' May 2 , 1912 .
448 Mr. F. W. Aston .
Influence of Nature of [ Aug. 23 , Hydrogen .
" A. B. E. F. F/ A. Mg 23 0 39 190 50 217 A1 23 0-41 170 66 287 Fe 34 0 *44 260 91 267 Cu 47 0*45 300 130 277 Zn 43 0*41 220 118 275 Ag 51 0*47 325 140 275 Sn 41 0*45 250 ' 120 294 pt 45 0 *42 270 120 268 Pb 51 0*50 320 166 310 C !
39 0*43 230 137 284 The above table gives the mean values of the constants of the two empirical equations worked out where possible for all the cathodes used .
In addition , in the case of oxygen , are given Di , the length of the dark space at unit current density and a pressure of one-tenth of a millimetre , and the potential under those conditions .
These are obtained by interpolation and so may be regarded as the most accurate of the numbers given .
It will be seen that the values of A , E , F , Di , Yh follow very approximately the same order as the cathode falls given by Skinner* and Mey.f F/ A is practically constant and very nearly the same for both oxygen and hydrogen .
The values of Vi and Th are exceedingly interesting and are plotted against each other on the accompanying diagram ( fig. 2 ) .
All the metals lie on a straight line through the origin except Sn , Pt , and Pb , which are about equal distances above it .
The value of B cannot be determined with much accuracy but shows no striking variation at all .
The edge error tends to make these values of B too low .
This remarkable constant , which represents the contraction of the dark space with increase of current density at the surface of the cathode , has already been shown to be almost unaffected by the pressure and the nature of the gas , } and is now proved by the results of this and the previous paper to be also practically independent of the nature and curvature of the cathode surface .
Turning to the influence of the nature of the cathode upon the dark space at constant C and P , its change of length may to a certain extent be accounted for by some such hypothesis as this:\#151 ; Let us suppose that under such unit conditions the density of the ionisation * ' Phil. Mag. , ' [ 6 ] , vol. 8 , p. 387 .
t ' Yer .
Deut .
Phys. Ges .
, ' 1903 , vol. 5 , p. 72 .
+ F. W. Aston and H. E. Watson , 'Boy .
Soc. Proc. , ' 1912 , A , vol. 86 , p. 168 .
Cathode on Length of Crookes Dark Space .
at the edge of the negative glow is the same for all cathodes .
It is quite probable that this density of ionisation is a function of ( 1 ) the number of cathode particles emitted from the cathode surface , ( 2 ) the distance through which they travel , Di , and ( 3 ) the total energy available from each , Vi .
How if any one of these variables becomes less , one or both of the others will probably become greater , so that as we pass along the series from metals which easily give off electrons to those which do so with more and more difficulty , we Sn 0 X '6 -8 1-0 1*2 1-4 1-6 Length of Dark Spa.ce in cms .
should expect an increase in the length of the dark space , and the fall of potential across it , as is actually the case .
This rough hypothesis will account for the supply of positive ions at the edge of the negative glow , but when we come to consider more closely their transport across the dark space to the cathode the most formidable difficulties are met with .
Thus the author has shown , * by a method which seems very free from objection , that for an aluminium cathode the density of positive electrification p is practically constant throughout the dark space .
It is equal to V/ 27tD2 , and the electric force at the surface of the cathode 2Y/ D. * F. W. Aston , ' Roy .
Soc. Proc. , ' 1911 , A , vol. 84 , p. 526 .
VOL. LXXXVII.\#151 ; A. 2 I 450 Influence of Cathode on Length of Crookes Dark Space .
Taking the actual case of aluminium at P = 0T mm. , D = 1-035 cm .
, V = 492 volts , and C = 01 milliampere or 3 x 105 E.S.U. , works out to be 0-244 E.S.U. per c.c. , so that even if we give the positive ions the velocity-necessary for the carriage of the -whole of the current passing through the tube this will only amount to 1*23 x 106 cm./ sec. , a speed nearly ten times too small for the mobility worked out for the pressure and electric force given above .
It is very unlikely that such a large discrepancy can be met by supposing that the ionic mobilities at such low pressures do not follow the ordinary laws , for Mr. Todd 's most recent measurements of mobilities at pressure of this order show an increase rather than a* decrease in the expression velocity x pressure/ electric force for positive ions .
A more reasonable explanation is supplied by the hypothesis suggested by Skinner* that many of the positive ions rebound from the cathode without giving up their charges , but in this case the electric force should increase much more rapidly than it does close to the cathode .
A moment 's consideration will show that surface layers of positive electrification advocated by Skinner and lately by Westphal , t even if they exist , are of no assistance , as in this case the actual electric force and its rate of change ( 47 were directly measured within a millimetre of the cathode .
Again if the distribution of electric force is the same for other metals as for aluminium the values of Vi , Di , given above show a constant field at the cathode surface with a value of p continually diminishing as we go from Mg to Ag , so that in order to supply the deficiency of current we should be forced to suppose the percentage of this carried by the negative ions greater in the case of silver than in that of magnesium .
This seems unlikely , and at first sight to be in direct contradiction to our original hypothesis ; but the possibility must not be ignored that the electrons emitted by such a metal as silver may start off with too high a velocity to be efficient ionisers .
The attack on these and other difficulties would be greatly assisted by even rough direct comparisons between the current carried up to the cathode by positive ions and that carried away from it by negative ones .
Experiments have now been commenced with perforated cathodes in an attempt to solve this problem .
Summary of Results .
( 1 ) The relations between the values of pressure , voltage , current , and the length of the dark space are determined for plane cathodes of many different * 1 Phil. Mag.,5 1902 , [ 6 ] , vol. 4 , p. 490 .
t ' Yerhand .
d. Deut .
Phys. Ges.,5 1210 , vol. 12 .
Emissivity of Solid and Liquid Gold at High Temperatures .
451 materials , and found to satisfy the same form of equations as those previously given for aluminium , the constants varying considerably .
( 2 ) Roughness of the cathode surface does not appear to affect the discharge , if the dimensions of the irregularities are small compared with the length of the dark space .
( 3 ) The length of the dark space is shown , in the cases examined , to be greatest for silver and least for magnesium , the metals following the same order as in the case of the cathode fall .
( 4 ) The rate of change of length of the dark space with change of current density at the surface of the cathode seems much the same for all cathodes .
( 5 ) Difficulties in the way of arriving at a satisfactory explanation of these and other data connected with the dark space are indicated and shortly discussed .
/ In conclusion the author wishes to express his best thanks to Prof. Sir J. J. Thomson for his kind interest and advice during the investigations described in this and in the previous paper .
A Speetro-photometric Comparison of the Emissivity of Solid and Liquid Gold at High Temperatures that of a Full Radiator .
By C. M. Stubs , M.A. , 1851 Exhibition Scholar , University of New Zealand , and E. B. R. Prideaux , M.A. , D.Sc .
( Communicated by Prof. F. G. Donnan , F.R.S. Received June 21 , 1912 .
) ( From the Muspratt Laboratory , University of Liverpool .
) Introductory .
It is well known that copper and gold emit greenish or bluish light at high temperatures .
KirchhofFs radiation law obviously suggests a connection between this selective emissivity and the selective reflectivity of these metals at ordinary temperatures , which gives rise to their colour .
According to that law the emissivity E ' of a surface at a temperature T is connected with its absorptivity A , by the relation * E'/ E = A , where E is the emissivity of a full radiator or " black body " at the same temperature T. This relation holds for each particular wave-length .
The ratio E'/ E may be conveniently called , and will be referred to hereafter in

