Active Nitrogen: A New Theory

BY

s. ^K.

^MITRA.

The Indian Association fo-! the Culti,::;ion of Science honoured me by awarding the Joykissen Mookerjee Gold Medal for the year 1945. According to the terms of the award the recipient is required to deliver a course of lectures on a scientific subject and to submit to the A<;sociation a copy of the lectures delivered in a form suitable for publication as a Special l\Iemoir. I selected as my subject Active Nitrogen as I had recently had occasion to study it, having proposed a new theory of this elusive substance. I also took this opportunity to write out the address in the form of a monograph, as I found th.at in spite of the wide interest of physicists and chemists in the subject, the one and the only report on it that has appeared so far is that by H 0 Kneser m 1929. (Erge-bnisse der Exakten N aturwissenschaften, Vol. 8, p 229). The report is an excellent one but is unfortunately outdated. I therefore thought that a connected account .jjJ the present state of our lffiowledge of Active Nitrogen would be helpful to the workers on the subject. Part I of the monograph has been prepared with t~is object in ''iew. In Part IT the new theory is discussed and is applied to explain the variou~

properties and phenomena of Active Nitrogen.

It it; a pleasure to record my thanks to 1\Ir. ,J. S. Chatterjee, M Sc., Ghose Research Scholar, who assisted me in the preparation of the monograph m many ways. But for his enthusiastic co- operation the timely completion of the monograph would have been very difficult. Mr. ChatterJee will shortly be leaving for England as a Government of India scholar; this will greatly interfere with my plan of carrying out with his assistance certain experiments for further verification of the theory.

I also take this opportunity of thanking the authorities of the Indian Association for the Cultivation of Science for tlw promptness with which they arranged "the publication of the monograph.

September, 1945, Wireless Laboratory,

University Coliege of Science, 92, Upper Circular Road, Calcutta.

S.K.M.

---~---

\

THE ... 'fi ... ., ... . Zfl, ' ll.ztl'3trM, ti~,A· ~lA ^lltJ" I

0 CUTT•

----

-..~;...---- CONTENTS

PART £

lACS

111111111111111111111111111111

6587

PAGE

1. Introduetion 3

2. Experimental "Methorls of Pt·odu<'ing the Afterglow 5 3. Life of tlw Aft<'q!low; the Wall Effe<'t 7

-t Layer Fo1·mation 8

!1. Laws of Dec· a,\- of the Afterglow 1-l

6. Pressure Effeets 18

7. Temperature Effeet 21

8- Spectrum of the Aftet·glow 22

9. Excitation of Spec·tnt 2-l

10. Chemical Activity 26

11. Phosphorescence 29

12. Dark )fodification 30

13. Reproduction of Kight Sky and

..

Au rorul Spectra in th€' Afterglow

H. Ionisation in the (:lowing Cas 1!i. Energr :\Ieasurements

16. Current Theories 17- Summary

31 33 38 41 44

1. Introduction

I

THE NEW THEORY

(

2. Long Life of the Afterglow 3. Spectrum of the Afterglow 4. Layer Formation

5. Rate of Decay of the Afterglow; Effects of Temperature and Pressure V ar1ations

6. Energy

7. Chemical Action and Spectroscopic Excitation 8. Ionisation

9. Concluding Remarks References

PAG1'~

49 53 55 57

59 60 63 64 68 71

Indian Association fol- the Cultivation oi. Science

ACTIVE NITROGEN-A NEW THEORY

BY

S. K MITRA, D. Sc.

Sir RMhbehary Gl10se Professor of Physics, Uni-vers1ty of Calcutta.

CALCUTTA

1945

---~~---

1

THE .... tll.nmiM tl~tA• ~lA.

^"QJU.

tU" I

Z1U, tJowbszll 8t!Wf ·

1 _

OAL. CUITA. '

-

Printed 'by J. N. Bose, at the Indian Press I-'td., Calcutta and Published by the Indian Association for the Cultivation of Science, 210, Bowbazar Street,

Cale.utta.

PROPERTIES AND PHEN"OMENA OF ACTIVE

NITROGEN

It has long been known that the glow in a tube conh~ining nitro- gen at low pressure and subjected to strong electric discharge, per- sists for a considerable time after the exciting Yoltage has been cut off.

This so-called afterglow o£ nitrogen was first noticed by "'arburg in 1884 and systematically studied by P. Lewis in 1900. The phenomenon has since been in':\·estig-ated in its various aspects, e.g., duration of the afterglow, its spectrum, its laws of deca,'l·, special methods of excitation, energy of the g·lmYing g-as, !.'tc., hy many workers all over tlw world. Sperial mention should be made in this connection of the work of Lord Ra,'l·leigh. His first contribution to the subject appeared in l!Hl and he is still active!? eng-aged in its study. His recent series of papers (1 !}:3;}-42) are a delight to :any the01·i-;t as they proYide for the first time sound material for building a theory. Rayleigh was a I so the first to rrcog- mse and study the <'hemirally actiw natnrp of the glowing gas and !.!,ave it the name · · ActjYe Xitrog-l'll · '. As a matter of fact. a report on ~\ctiw Xitrogen is, in its major part, a report on the work of Rayleigh. .Amon-gst other workers who have made systematic study of the subject, mention may be made of E. J. B. Willey in England and of B. Lewis in the u.S.A. Other notable contributors are Kaplan in the U.S.A., Cario, Franck, Herzberg, Kneser in Germany, Okubo and Hamada in .Japan. Saha, Kichlu and ,Joshi in India.

Various theories have been proposed from time to time regarct·

ing the mature of active nitrogen. Cnfortunately none of the theories offered till now has prowd satisfactor~- to any degrel'. The author of this monogl"aph has recently proposed a theory which offers a simple explanation of the characteristic propeJ·ties of active nitrogen and the phenomena associated with it. In the followin·g pages a review \Vill first be made of the present state of our know- ledge of a(•tive nitrogen and of the cmrent theories. The theory proposed by the author will then be discussed.

The chronicling of facts regarding- actiYe nitrogen which may be considered as established beyond doubt, is n·ot alwaysean easy task. Any one attempting to study the literature on the subject is not ml'ly struck by the large amount of experimental work done but, at the same time, is greatly <'onfust•d h,'l· the divergent rcs.ults

and conflicting statements of the d~· 'erent authors and also, not infrequently, of the same author on .ifferent occasion'S.

A typical and perhaps the best i lustrative example of this may be cited here. This is ·~n ₍ the effect of slight admixture of 0₍ 2 in the nitrogen employed. While Strutt (Rayleigh) (1915) in London firmly held that oxygen is not necessary for the production of the afterglow, Tiede and Domcke ( 1913) in Germany held equall,v firmly that slight admixture is essential for its production. Th,, controversy was only settled when Tiede and Domeke brought over their apparatus from Germany to England to demonstrate their point to Strutt. The result of this is perhaps best described in the words of Strutt, ''In the controversy which has been reviewed it was maintained on the one side, that pure nitrogen would give the full effect, and, on the other, that the presence of oxygen was essential. It is now seen that as in so marty scientific controversies neither side was entirely right. Almost an.y contaminaton with the exception of argon and helium, increases the yield of actiw nitrogen, as judged by the intensity of the nitrogen afterglow",

There was a somewhat similar controversy over the effect of the introduction of neutral nitrogen into the glowing gas-a qucs tion of importance in the theory of active Il'itrogen. While one group of workers (Kneser, 1928) maintained that addition or

neutral N2 in the afterglow increased the intensity of the glow and diminished the life, another group reported that addition of the neutral gas had no effect (Bonhoeffer and Kamin'Sky, 1927). The question has been settled only recently when Rayleigh (1940) by careful experiments showed that the former was the correct view.

Irrstances can be multiplied but from those given above it is evident that it is not always easy to make categorical statements on experimental results which may be regarded as e&""tablishen beyond doubt. The reader will presently see that the contradictory results and statements are due to the fact that the pro- perties exhiuited by the glowing gas are extremel.v sensi- ti\'e to the physical conditions under which the gas is excited, e.g., the purity of the gas employed, the condition of the walls of the containing vessel, the mode of excitation, etc. Attempt will nevertheless be made to include in the sun-ey all the experi- ments and experimental results which, in the opinion of the author', are of d furtdamental character, and as such provide basis for the formulation of the theory.

2. EXPERIMENTAL MEfHODS F(lil PRODUCING THE AFTERGLOW,

The usual method for producing the afterglow is by the so- called condensed discharge. A typical arrangement for the purpose is shown in Fig. 1 (left). The electrodes used are long to allow for the dissipation of heat. Electrodeless discharge may also be used, the advantage being that no metal comes in contact with the gas in the discharge, Fig. 1 (right). In this case the discharge consists of highly damped electric oscillations. One may also use continuous

Fig. 1. 'Experimental arrangement for prOducing nitrogen afterglow; (a) by condensed discharge, (b) by eleetrodeless discharge. Typical values:

Secondary voltage-20,000 volts; con.-l.enser-0'01 p.F; pressure-Of the order of a millimetre.

oscillations generated br a transmitting valve. The afterglow may also be produeed by ordinar;•: dirPet diseharl};e. Kaploo (1932) has prepared special tubes by prolonged discharge-running to several days-in which the afterglow produced by ordinary discharge is quite strong. According to ,Joshi and Purushotham (1939) too much stress has undul~· been given on excitation by condensed discharge :Jior produetion of active nitrogen. With simple H. T. they succeeded in getting sufficient aetive nitrogen.

Active nitrogen can also be produced by arc di~~harge. Thus.

in one of the experiments of Constantinides ( 1927) the following arrangement was used: ''A steady arc was maintained with approxi-

mately :JO Yolt-; "·hile the gas was allO\/ed to flow past into the cylin- 'lrieal eleetrode. The characteristic fellow glow of actiYe nitrogen (•ould he obserwd in th~ neighbourhood of the arc".

Kenty and Turner ~1928) obtdined actiYe nitrogen by the hom h<mlmcnt of nitrogen gas with electrons of energy greater than 16.:3 eY.

It ma~· lw mentioned here that all attempts to condense the actiYe substance by liquid air met with negative result.

The properties of the afterglow may be studied either with the

•liseharge running or with discharge stopped. The fir<>t method is a1loptt>d when the life of the glow is short or when long continued ob-;crYations are to be made. In this ease the glowing gas is conti- nuously pumped out of the diseharge tube and can be studied in a Yessel through which the glowing ~as flows on its way to the t'xhaust. The delay between the aetual discharge and the afterglow in the obsenation vessel depend'! upon the rate of the flow of the

~lowing- ~a"> and on the distance of the former from the discharge tuhe.

Tlw afterglow may also be studied, as in the ease of the study 1)f the Jaws of deeay, b~· cutt~ng off the discharge and the pump.

lt is of course necessary in this case that the life of the glow be long- (see See. :J).

The true nitrogen afterglow-also called the Lewis-Rayleigh

~low-ha-; a ri<'h golden yellow eolour (see Sec. 8). If impurities are prl'-;ent in the nitrogen u-;ed, the colour of the afterglow may be (lift'erent. Thus, oxygen, a common impurity, if present in large

quantit~·, produee<> a deep green afterglow said to be due to reaction between :\'02 and 02 (Rayleigh, Hll2a). Oxygen in smaller quantity imparts a blue eolour to the g-low due tl) the {3 hands of NO. Hydro- (•arbons derived from tap g-rea'"ie give a violet colouring 11ue to CN hands. unless preeantions are taken the~e colours may ma"k the true colour uf the afterg·low.

As mentioned in the introdnetion the i11tensity, as also the life of the glow, is markedly dependent ^OI! the nature of the walls of the

·eontaining wssel and on the purity of the gas employed. The larger the wssl'l containing- the g·lowing gas the greater i'l it'l life.

Rayleigh ( Hl40) in some of his expPriments usPd vessels of 19 litrt'>

<·apacit~-, while Angerer ( 1921 ) U'\ed wssels of 52 litres rapacity (diameter 46.5 em.). The life i-; also dependent on pressml:'. It is longer at low pressures. The decay time. howevt>r. ··rearhes ·a <'OilS·

tant limiting value as the pressure is indefinitPly (liminished. '' For

example, "b!'tween 0'05 mm \1nd 0'10 mm there is little change in the time of decay over a gi \·e1. intensity" ( Ra~·lei~h, 1 !J!l;), see also Kneser, 1H:?R). It has also been found tflat slight admixtm~e of impurities (up to about 0'1%) of sontc !'lectrOJwgative element enhance-> the glow. The significance of these results is discussed in the following Section.

3. LIFE OF THE AFTERGLOW: THE WALL EFFECT The most important f<wt regarding- the deea~· of the aiter~low

that has emerged out of varied experiments and ohst'ITations is that the artive substance is destroyed t)artl~· in the volume of the gas-. in (•mn·se oi which the glow is emitted, and pat·tl~- on. the surface of 1 he gla.ss wall, which is a glow less proress. A dirert t•vitl<.'ncc of this is that, otlwr rontlitions remaining the same. the larg!'r the size of the n•ssel the longer is the Jifl' of tlte g-low. This is easil~- cx- plainl'cl. With the inrrease in the size of tlw Yesscl the ratio hetwern the surfare arra and tlw volt1nw derr(•ast>s and a smaller proportion of the a<'tiw substance i~ llestroye«l 011 tlw wall b~- the g-low less prorrss.

Hegarding the effect of the wall it has been fonn(l that the rate of deeay of the glow might be areell•rated or l'etarded hy spc<'ial treatment of the wall. The following results are well established by expt>rimrnt.

JJestructil'e of the Glow.-(1) \Y a lis of nssels which have been strongly heatPd for several hours. (2) Coating the glass surface with "poisons'', e.g., .\piezon oil. (3) Admixture of hydrogen.

Protectiee of the Olow.-(1) Coating the g-lass surface with speeial snhstances c.{f., silver film, metaphosphorir acid. (2) For- mation of adsorbed layers on glass surhtee. This is produced when the nitrog;ru used contains impurities of eleetroneg;ath-~ elements, the most important of whirh is oxygen. (3) Conditioning the wall by prolonged <liseharge. This also produces adsorlwd layer on the surfare. \Ye now des<'ribe some of the ex)wrinwnts which lun·e leJ to these <'onclusious.

Heat 1'reatmenf .-Henerally, a glass vessel which has been

<'leaned with rhemicals (say <'hromic aeid) maintains the gl8w for a moderately long time. If howcwr the vessel be cleaned by lwating;

strongly to 400°C or above then the clean wall of sn<'h a wssel is strongly destructive of afterglow. The experimental arrangement

shown in Fig. 2 due to Rayleigh (1.$42) shows the effect clearly.

A and B are tubes of Pyrex glass. A/is previously heated by a resis- tanl.le furnace for aboqt an hour to a temperature of 500°C the pump acting all the whillf. On co~ling, when the glowing gas is admitted it is found that while the glow in B is bright that in A

is scarcely \"isible. Evidently active nitrogen is being destroyed by the wall as fast as it is entering. The gl0w in A is gradually res- tored if the supply of active nitrogen is kept up for a long time.

Special Coating of the Wall.-The waH of the containing vessel may be rendered favourable for long life by coating it with suitable substances. B. Lewis ( 1929) and Willey (1930b) report that paraffin formed a protective layer. Rayleigh (1935) found that coating the inside surface witp. metaphosphoric acid or sulphuric acid has a marked effect on the life. With a. vessel of 12.8 litre (29 em. diameter) capacity, coating the inside with metaphosphoric acid and with nitrogen at a pressure of 0"35 mm., Rayleigh was able to prolong the life to more than 5! hours. Apiezon oil on the other hand has a destructive effect.

Purushotham (1939) found that coating the wall with silver film has the maximum effect in prolonging the life of the afterglow. For other materials tested, the influence is in the following descending order: silver film, metaphosphoric acid, sulphuric acid, cleaning with chromic acid, arsenic oxide, thorium chloride.

We will now discuss in the following Section the formation of adsorbed laye.rs on the glass surface which is protective of the glow.

4. LAYER FORMATION

The question of the formation of a protective layer on the suu~

face ros<: early in the study of the afterglow when it was discovered that slight admixture of some electronegative element as impurity (in particular oxygen) had the effect of increasing the intensity and the duration of the afterglow. As a matter of fact, all prominent

investigators on active nitrol\;n have, at one time or other, been forced to the conclusion that -J,1der certain experimental conditions, the walls of the afterglow vessel are cove.-ed by a layer of ·gas which prevents the destruction• of the active substance on the sur- face. 'Ve may quote here some of the remarks made:

Rayleigh (1942) while discussing the effect of slight admixture of oxygen in increasing the glow, says, ''To be more specific, it seems natural to assume that the admission of a trace of oxygen, of the order of a fraction of one per cent of the nitrogen, leads to the building up of a layer on the glass which has a favourable effect".

Willey (1930a) in his studies on the rate of decay of the glow when different gases, are admitted into the afterglow tube, remarks,

"Unless the walls of the vessel concerned are "poisoned" by foreign gases, the recombination process is non-luminous and occurs as a surface reaction''.

''As the quantity of gas added is increased, the adsorbed film becomes thicker and more extensive, so that finally a point is reached where the active nitrogen entities can reach the metal only upo11 rare occasions and are compelled instead to suffer deactivation as a luminiferous process in the gas pha~" (Willey and Stringfellow, 1932).

B. Lewis (1929) in his study on the influence of the. surface on the afterglow remarks, ''The nitrogen afterglow is made observable only after the surface of the vessel has been effectively poisoned by the adsorption of a layer of gas''.

Kaplan (1932) in course of his experiments. on the preparation of special tubes for production of auroral spectrum in the after- glow, says, "The long running of the tube undoubtedly conditioned the surface of the tube in such a way as to allow it to adsorb nitro- gen atoms.'' Again, in discussing the action of hydrogen as a 'poison' he ( 1935) remarks, ''Hydrogen may inhibit the afterglow by being adsorbed on the walls of the tube, thus preventing the adsorption of nitrogen atoms and the subsequent recombination to form meta-stable atoms".

It was not at first clear if the impurity took part in the mecha- nism of the production of the afterglow or, if it acted on the stlf- face of glass, so conditioning it as to prevent the de-activation of the active substance on the wall. Subsequent investigations 4ilhowed that the latter was the case.

\\T e describe below some experiments which led to this conclu- sion.

(1) Willey's E:rpe?·iments.-Wil)cy (1930a,b) designed ex- periments to study how the rate of {decay of the afterglow varied whe.n impurities of difftrent kinds were introduced in the nitrogen used. He found that witl-. the gradtwtl introduction of the impurity the rate of decay approached to that corresponding to bimolecular '

reaction, a reaction which may rpasonably be expected to proceed in the volume. "'hen pure g·as wa~ used the rate strongly departed from the eune for volume reaction. WilleY concluded from this

. .

that the impurity formed protertiw ht~·er on the glass surfaee and c-ompelled the de-actin1tion to proeeed in the volume of the gas.

~0.---~

+

a~~~~~~,~o~,~~~,~o~z~~~l~o~~~~~

Oact

Fig. 3. Illustrating how the 1ate of decay of thP afterglow changes "hen the amount of impurity in the gas is varied. D =distance of the point of observation from the discharge tube. 'fhe ordinate is so chosen that if the decaY is due to bimolecular reaction in the volume (t vmfr.lg as 1/\/I) the curve \\Ould be a straight lm'l>. Amount of 1mpurity: A-'03o/r, B-'07<fr, C-0'1%, D-1'3%. (AftPr Willl'y).

Figure 3 shows the decay curves us obtained by ·willey in some of his experiments. The ordinate seale is proportional to

1/yl

where I is the intensity of the glow. With this scale the intensit~y

time curw will be a straight line if thP reaction process be bimole-

t"

cular in volume (see Sec. 5). If, on the other hand, the decay be on th£> surfaec, the rate will he that corresponding to monomolecular

reaetion and the 1/\II-t graph will be strongly curved.

It is seen that lrhen tlle ~-npurities m·c snw11 ( eUlT<'S A and B) the graph is strong]~· emTed &\1owing· that the deea.'' is not proceed- ing in the Yolume. Introduction of impuritJes in larger quantities, conditions the wall and the c1ene heeonl{s substantial!.'- a straight line (eurn• C). "'ith still further inc•rease. the impuritr hegins to react with the aetiYe substanc-e, inereasing its rate of decay and the graph therefore again become~ strongl.'- eurwd (rune D).

The nitrogen used by Willey was freed from 0₉ by passing it over hot copper ( 600•C) and containe<l a" impuritie~ 0'3% Argon and 0'01'/c other gase,.

When l•xrited it gave ver~· feeble glow on account of its purlity. The curves in tht>~e experiment~ were obtained as folio"~- '!'he inten~it~· of the afterglo\\

was measured at different distances (D) along tht> tubl•. Since the glowing gas wa;;; flowing to the exhaust at a constant rate the distance~ were obviously proportional to the time. For moosuring the inten~ity two S{X'cial caesium

~lis wt>n> H"ed in parallel. The~P wert> phtrerl in sel'ies with a Tinsley moving

<'Oil galvanometer (~eusitivit~· 5X 10-"' amp.) an1l n 150-volt hnttery.

(2) Rayleigh's E;rzm·iment.-Perhaps the most eonvineing test that the impurit~', nt least so far as ox,\·gen as an impuritr is con- eerned, acts on the surfac-e is furnished L,\- CX\pei4iments of Rayleigh

(Hl42). Ra~·leigh madf:' wrr careful-experiment to study how the intensity of the afterglow Yaries when ox~·g-en in measured quantit~·

is mixed "·ith nitrogen hefore its passag-e into the. dischargP. It was found that thr time requirell for the btahlishment of the glow to its full ntlne is mnch longer than the time required for the addetl oxygen to diffuse and fill entirely the obsenation Yessel. SimilarlJ·, when the ox~·gen supply was shut off, the glow pprsisted long after the oxygen in the observed Yessel had been washed out. In a parti- cular expriment, the time required for the oxygen to completely fill the obsenation tube was 0.91 see., while the time for the establish- ment of the full intensit.'· was

9l

seconds (maximum mean value of several runs). Similarly, the time required for the oxygen to be completely washed out was 0.91 seconds. whereas, the increased glow persisted for 61 seconds (maximum mean of several runs). Em- ploying a tube the mtll of whieh had been "poisoned'' b,\· heating and, as a consequence, the aftprglow in which was very feeble, the intensity was found to inc·rease 32 times 'vhen oxygen in amount 1.4% of nitrogen, was introduced.

From these experiments Rayleigh concluded that the oxygen present (of the order of a fraction of o.ne per eent of nitrogen)

"leads to the building up of a layer on the glass which has a

•

favourable effect'· on the life of afterglow.

( 3) /{a plan's Experiments.-Conditioning of the n·all by pro- longed discharge. Kaplan ( 1932) \vas able to obtain bright after-

glow with ordinary discharge in a tul}4 filled ·with nitrogen together with a small amount of oxygen (see Sfe. 13). He observed that when the ·tube was run steaaily for several days oxygen was gradually cleaned up and the green cl.ntinuous glow due to oxygen and nitrogen mixture gradually disappeared. At the same time the afterglow due to active nitrogen appeared with increased intensity. Accord- ing to• Kaplan the long running of the tube conditioned the surface of the tube in such a way as to allow it to adsorb nitrogen atoms.

( 4) K enty anJd Turner's Experiments.-The most direct proof that adsorbed layer is formed in the presence of active nitrogen is furnished by the experiments of Kenty and Turner (1928). These investigators first noticed that electron emission from a hot tungsten filament is diminished in presence of active nitrogen: They subse- quently found that resistance of a fine hot ( 400°C) tungsten filament is considerably lowered if active nitrogen flows past it. Both these effects were traced to a lowering of the temperature of the filament due to the formation of a molecular (or atomic) layer on the fila- ment which caused a part of 'the heat to be conducted away by the surrounding gas. To test the formation of the nitrogen layer on tungsten, Kenty and Turner ;performed the following experiment.

A bulb as shown in Fig. 4 contai:!ring a number of circular

G ^F

I

F1g. 4. Kenty and Turnel"'s apparatus to determine the amount of mtrogen adsorbed by tungsten surface m the presence of act1ve rotrogen.

tungsten discs was taken. (The discs were 42 in number and each of 8 mm. diameter). Nitrogen activated by discharge

between the electrodes E formed adsorbed films on the, tungsten discs. G is a ealibrated hot-~re pressur~ gauge. F is the test fila- ment of tungsten, the lowering of the resistance of which indicates the presence of active nitrogen. The tot\1 ~olume ^{of the} appar~tus was 18 em. ³ while the area of the exposed tungsten surfaces was 42 em.z. The whole apparatus was first baked out at 440°C and also spark passed between the electrodes to drive out all adsorbed gases. Pure nitrogen was then admitted to a pressure of 1.9 mm.

and the volume shut off. Condensed discharge was passed between the eleetrodes and, at the same time, the tungsten discs were kept red hot by intermittent use of induction furnace. After the experi- ment, when the tube was cooled, it was found that the pressure had dropped to 1.4 mm. In order to recover the adsorbed gas the tube was evacuated, shut off, and heated at 400°C for 15 minutes in furnace and the filament also flashed. Approximately .05 mm. of gas was thus recovered. By further treatment a total of 0'2 mm.

of gas was finally recovered. Assuming that this is the quantity adsorbed by tungsten discs it is of the•order of magnitude necessary to rover tungsten surface with a single layer of atom with one atom for each square Angstrom.

It should be mentioned that Kenty and Turner did not find any adsorbed nitrogen from the glass walls of the apparatus.. This is of course not to be expected. Firstly, because the glass had been heated to a high temperature before the activation of nitrogen and it is known from the experiments of Herzberg ( 1928a), Rayleigh ( 1911) and others, that heating of the glass vessel renders it un- favourable for the afterglow (see Sec. 3). This means that the adsorbed layer which prevents the destruction of the active subs- tance on the surface is not formed on the surface of glass whi~h has been strongly heated. Secondly, according to the experiments of Rayleigh and of Willey already quoted the laye11 on glass surface persists only as long as oxygen as impurity is present in the gas.

Kenty and Turner made another interesting obser;,ation which has a bearing on the favourable effect which oxygen has on layer formation. They observed that if 02 is present as an impurity in the nitrogen used, a slow and steady lowering of the resistance

o~ the filament occurs with time even when there is no activation.

·I.n

such case if the gas is bombarded with electrons of en~rgy of approximately 10 volts the layer formation quickens and there is a sudden drop in the resistance. If, however, sufficient oXi,ygen is present, spontaneous lowering of the resistance is so rapid that no

further lowering is observed when bombarded bv electrons and the passage of active nitrogen over the t(.ngsten: fila~ent has no cooling effet:lt; on the contrary it may produce a slight warming. This latter effect had been observed l>y Willey, (1927a) in his experiments on heating of tungsten by aclive nitrogen.

It would be interesting to note here the comparative efficiency of the different substances in promoting or hindering the layer forma- tion. Rayleigh (1915) in one of his earlier investigations carried out tests on various substances and listed them as below in order of efficiency.

Hydrogen sulphide Water

Carbon dioxide

Oxygen Mercury Chlorine

Carbon monoxide Hydrogen

Acetylene Argon

Ethylene Helium

Methane Nitrogen

According to tests carried out by Willey (1930b) the efficiencies of C02, N20, N2 and CO are in the order mentioned.

The case of hydrogen needs special mention; instead of merely baing a diluent, it has an adverse effect on thf.' wall. We have, amongst others, the following significant observations,

(1) Knauss (1928) found that hydrogen quenches the after- glow.

(2) Kaplan (1935) was able to obtain afterglow in uncon- densed discharge (in his special tubes for reproduc- tion of auroral spectrum; Sec. 13) only by removing the slightest trace of hydrogen.

(3) Kenty and Turner (1928) found that a trace of hydro- gen quickly cleans off the nitrogen layer formed on tungsten filament and also prevents any further for- ' mation of the nitrogen layer. But curiously enough

hydrogen has no such effect if the layer formed is due to oxygen.

5. LAWS OF DECAY OF THE AFTERGLOW

The rate at which the intensity of the afterglow decreases indi- cates the type of the reaction which causes the destruction of the active substance and thus throws light on its nature. As already

mentioned the destruction pr~.('eeds partly by reaction in the volume- and partly by reaction on the walls of containing vessel-the latter being a glow less prO<'eS!\. We discuss belo\f the theocetical reaction rates for the two processes (ltinshelwooA, 1933).

Decay in the t•olume.-The reaction in the volume may be mono- molecular, bimolecular, termolecular or of higher order, according as one, two, three or more of the active molecules take part in each act of emission of the afterglow.

For monomolecular reaction the active particle reverts to the normal state either spontaneously, or as a result of collision with a neutral particle. In either caM, if N A. be the concentration of the active particles at any instant then, assuming that each act of rever- sion is aecompanied by emission, the intensity of the glow is given

b~· I= KN A., where K is a constant. But, this is also the rate of disappearance of the active particles. Hence,

Integratin~. - K I log N A.= t +constant.

Or, since I is proportional to N A., 1jlog I varies as t.

Thus, for monomolecular reaction, the 1/log I-t curve is a straight line.

If the decay process involves reaction between two particles of the active substance, then the rate will be that for bimolecular re- action, the intensity at arry instant being proportional to N A.^2•

Thus I= aN.~_ ^2, where a is the coefficient of recombination and is a constant.

But this also is the rate with which the active particles are dis- appearing. Hence,

al\;A. ²

= - d~

^{A. or,}

dt '

•Integrating ¹

!_ =

^t

+

constant.

a NA.

Making appropriate substitutions from the first relation, we have 1/y'l varying as t i.e., in this case 1/yl-t curve is a straight line. This is the law for bimolecular reaction.

If the reversion to normal state if due to reaction between three actiye centres then,

I

=

aN ,..^3, where a i~. a constant

_ dNA.

- - - -

_dt

Proceeding similarly as above we find that in this case 1ji213 varies as t, i.e., the curve

1/F'a

- t is a straight line. This is the law for termolecular reaction.

Generally for a reaction in which x particles (x>1) take part in each reaction process, the ~-₁ - t curve is a straight line.

I X

Decay on the W all.-If destruction on the wall be the predomi- nant process then the rate of decay of the glow will be that cor- responding to the monomolecular reaction. This is easily seen as follows:

Let N ^A be the concentration of the active substance at any instant. Then the rate at which the active particles w1ll be dis- appearing per second is proportional to N ^A and to the area of the

"poisoned" glass surface which is constant. Hence, dN4.

- - - = K NA where K is a constant.

dt

On integrating, -

~

log N ^A = t

+

constant,

Now the intensity I of the glow at any instant is proportional to N ^{A ..} where x = 1, 2, 3 etc., according to the order of the reaction.

Substituting I for NA in log N ^A we find that- log I varies as t i.e., when the wall effect predominates the

1/

log I-t curve is a straight line. (The active substance will also be destroyed in course of the emission of the glow, but its rate, by assumption, is negligible com- pared to that of disappearance on the surface).

The laws of decay have been studied by many workel"S. Of these, the latest and the most precise are those by Rayleigh ( 1935).

He used vessels of large diameter and, with specially coated bulbs, was able to prolong the life of the visible glow to several hours.

Rayleigh's experimental arrangement i'l shown in Fig. 5. The bulbs are of lead glass 29 em. diameter. For purposes of obserYa- tion the active nitrogen is produced in the bulb A and is admitted into the observation bulb B with specially treated wall. At the beginning the two bulbs are exhausted with cooled charcoal as far as possible. Stopcock D is then shut and a double dose of nitrogen

introduced into A giving a p.ressure of .07 mm. The nitrogen used, was from commercial cylinder, the oxygen being removed by stand-

•

Fii- 5. Arrangement of bulbs to study the laws of decay. Active nitrogen is produced in bulb A and is introduced in B where thf!

rate of decay is observed. By eoatim.g bulb B with special substances it is possible to prolong the life of the glow to several hours. (After Raylei,qh).

ing th~ gas over moist phosphorus. The gas was subsequently dried up by phosphorus pentoxide. For photometric measurement a sur- face coated with potassium uranyl sulphate, the faint glow from which could be'eonveniently matched with the afterglow, was used as the standard source. (For fuller details the original paper may be consulted.)

The rate of decay was studied for two cases. First when the inside of the afterglow vessel was moistened with Apiezon oil and secondly when it was coated with metaphosphoric acid. The decay was found to be much more rapid fpr the first than for the second case. To test the laws of decay the observed intensities were plotted on different scales.

In Fig. 6 the ordinates give stages at which the light intensity falls to half its original value. For instance, the intensity of the stage 5 is half of that of the stage 6; that for stage 4 is half of that for stage 5, and so forth. In this case if the law of decay is monomolecular the 1/log I - t curve will be a straight line. This is to test if the active substance is being des.

troyed on the surface. If the decay proceeds in the volume, with a reaction rate higher than monomolecular, then the half-value time will gradually increase. It is seen that curve b for the Ap§.ezon oil coated surface is nearly a straight line showing thai. the decay is proceeding on the surface of the vessel. Curve a for metaphos- phQric coated bulb, on the other hand., departs from straight line, the half-value time gradually increasing. This shows that in this case the destruction of the active substance producing the glow is flroceeding in the volume rather than on the wall of the bulb. It will also be noticed that at the beginning the two curves a~ nearly coincident. This may simply mean that initially, when the concen- tration is high, the number of reactions per second being propor- tional to that of the square of the active particles, (on the assump-

tion of a bimolecular reaction) the dec~ in the volume predommates even in the ease of the bulb with Apiezon coated wall.

Rayleigh also plotted the data for the metaphosphoric acid coated bulb with ordinate {~ales I 112 a~1d J-2!3 to test whether the law of decay in volume corresponds to bimolecular or to termolecular recation. Unfortunately both the curves deviated slightly from straight line and it wa.<~ difficult to Judge which corresponded more nearly to it. However, in one ef his later experiments usmg larger bulb (19 litre) Rayleigh (1940) optamed data which when plotted

•

!<'1g. 6 Rates of decay of the glov. when, (a) the mner v.all of

the

afterglow vessel ~~ coated WJth metaphosphonc ae1d, and (b) w1th Ap1ezon oil, The ordmates g:~ve the half value stag~>s. Curve b I!' stra1ght hne showmg that m th1s l'ase the active matenai 1s bemg destroyed on the wall (Aftrr Rayletgh).

2

o--~4--~,~~~--~.~-~~~~

•

Tl"'~ In ~c•nd•

F1g 7. Decay eurve drawn WJth data f.rom Rayle1gh's 1940 eJ.pen ment. W1th the ordmate scale lf\/1 the ~urve agrees excellently with a stra1ght lme. This shov.s that m the volume actwe 1nt.rogen 1s dell- troyed by b1mole.cular reactiOn.

as l j y l - t curve agree excellently with a straight line (Fig. 7).

This shows that the reaction order in the volume is bimolecular.

Chatterjee ( 1944-45a) has utilised these data to calculate the co- efficient of recombination and obtained the value 4X 1o-u cm³/sec.

6. PRESSURE EFFECTS

A question of great importance in the theory of active nitro- gen is the effect of the change of pressure on the glowing gas.

Three types of pressure effect might be distinguished ·

( 1) The glowing gas is compressed or rarefied as a whole, so that the concentrations of both, the active substance and the neuo- tral nitr-ogen, are simultaneously changed.

( 2) Keeping the partial pressure of th~ active substance cons- tant the pressure of the neutral molecules is increased by introduc- ing the latter into the gloWing gas.

A

8

D

Fig. 8. Appara.

tus for studying the effect of eom·

pre~sion on the glow inh!Jsity.

(After Rayleigh).

(3) Ke~ping the partial pressure of the 11eutral molecules constant that of active subs- tance is decreased.

Each of these three effects has been studied with great care and precision by Rayleigh (1940). His methods of measurements and the results obtained are briefly described.

( 1) In Fig. 8, A is the compression chamber.

The inlet for the glowing gas is B through which it enters from a store in a large bulb.

The piston C is made of India rubber and is coated with gummy semi-deliquesced phospho- ric acid. This latter also acts as lubricant. The piston rod is of glass passing through the baro- metric column and is controlled by the handle D. The inside o! the cylinder is coated with metaphosphoric acid to minimise wall effect.

In order to ob.serve the effect of compression the piston is first placed just above the inlet so as to isolate the gas in the cylinder. The inten- sity of the glow is observed. The piston is then quickly raised so as to compress the gas to half its volume and the increase in the glow is noted.

For observing the effect of rarefaction the arrangement in Fig. 9 is used. The two flasks A and B are of 1 litre capacity each and are moistened inside with strong sulphuric acid.

Active nitrogen is generated in B J:>y electrode- less discharge. The arrangement of the stop- cocks E and C to let the glowing gas diffuse in- to A and B is obvious. To study the effect of rarefaction the flask B is first exhausted with cooled charcoal and the glowing gas let into A.

C is then closed and the glowing gas is let into A through E. E is then closed and C is opened so that the gas passes into the vacuum B and the pressure is halved.

111111111111111111111111111111

6587

20 PROPERTIES AND PHENOMENA

The results of the experiments on compression and on rarefaction sho\\' that the intensity varies inversely as

I! "

cube of the volume. (It should be mentioned however that for the rare- faction experiment the diminution in intensity was found to be slightly less than that expected.)

(2) The effect, if any, of in- creasing the concentration of the inert gas keeping that of the radmt- ing centres constant had long been a subject of debate. In 1927 Bonhoeffer and Kaminsky reported negative result, though in 192S Kneser obtained a positive result.

Rayleigh even upto 1935, was not satisfied with the evidence of a

F1g. 9 Anangement to st~dy

the effect of reduction of pressure C!:J. the glo\\ m ten s1ty. (Afte1 Rayle,gh).

positive result. He, however,, made -..ery careful experiment<; m 1940 and showed conclusively that there was an increase in the glow when the concentration of neutral N2 is increa&ed, keepmg that of the active partides eonstant. In one of his experiments, Rayleigh (1940) used an uncoated bulb of 12 litre capacity carefully wa&hed with weak hydrofluoric acid and exhausted over P 2 0,. It was found that in this case the intensity increased 5 times when the nitrogen pressure was increased 5.1 times, the initial and the final pre&sures being 3"4X10-³ em. and 17"4X10-³ em. respectively. In another experiment a 19 litre globe coated with phosphoric acid was used. In this case also an increase m intensity, in nearly the same proportion as the increase in the gas pressure was observed. The initial and

(

final pressures in this case were 2"7 X 10-³ em. and 12"8 X 10-³ em. res- pectively. In both the&e experiments the intensity of the glow was very weak on account of the low pressure used. For higher pressure and bright glow the proportionality of the increase in glow and increase in pressure was however not mantained For afterglow ~

to 10 ti"mes brighter than that used in the above experiments, only a two-fold increase of light was obtained when the nitrogen pressure ' was increased 5 times. ^It is interesting to note that Kneser (1928)

also obtained similar result For a ten-fold incr·ease in pre&~urc

(from 5Xl0-³ mm. to 5Xl0-² mm.) ·he obtained only ·a threefold·1

increase in light intensity.

The experiments thus ^sho~that the a'terglow intensity increases with the pressure of the neutral N2. For low pressure the increase is almost proportional to the I?ressure of the neutral gas.

(3) In order to study how the glow intensity varies when the concentration of the active substance is decreased keeping that of the inert nitrogen constant the arrangement shown in Fig. 9 is used.

The whole system is first filled with nitrogen at low pressure of the order of 10-² em. Active nitrogen produced in D passes into A by diffusion, the stop-cock E being kept open. The discharge is stopped and after the temperature and pressure have settled down to uni- formity E is closed and C is opened so that the glow diffuses into B.

The concentration of active centres is thus halved but that of neutral molecules remains unchanged .• From a series of tests it was found that the intensity diminished about four-fold (actually the mean result was 3.4-fold).

It may thus be concluded that the glow intensity varies approxi- mately as the square of the concentration of the active molecules.

7. TEMPERATURE EFFECT

It has long been known that heating decreases the glow and cooling enhances it. In the original paper of P. Lewis (1900) we find the following remark: ''On admitting a slow stream of nitrogen the tube did not fluoresce at the heated portion although it did so on both sides.'' Similar effect of heating to weaken .the glow was observed by Strutt ( 1911) and by other workers, Okubo and Hamada (1933), Cario and Kaplan (1929).

The complementary effect, namely, increase of intensity by

~ooling has also been obsen·ed by several investigators. 'N e describe below the experiment carefully carried out by Rayleigh (1t?40) to study quantitatively the effect of temperature change. The experi- mental \'essel (Fig. 10) consists of a 300 c.c. capacity bulb with two legs 10 . em. long and 1.5 em. internal diameter. This \'essel is

, eonnected to a 19 litre globe and the whole apparatus is filled with glowing gas. ODi' of the legs is kept at room

F1g. 10 Expenmental ves ..el to study the effeet of temperature vanahon on the mtenslty of the glow. One of the legs IS kept at room temperature and the other IS 1mmer~ed 1D hqu1ds at dl:ffer<ut temperatures

(After Rayleigh)

temperature while the other is heated or

1 ooled by 'immersing, about 5 em. of its length, in hquids at d1fferent tempera- tures. The temperatures used were 100°C (boding water), -78"2°C (solid C02 in alcohol), -180°C (liquid air) and room temperature. Since the pres- 'lure in the observation ves'lel (being connected to the big 19 litre globe) re- mains practically constant, there will be change in concentration due to change of temperature in the observation leg. Due to this alone there w1ll be a change in the bght inten'lity. The effect due to this can however be taken into account by assum- ing 'the light mtensity to vary as the cube of concentration.

W~en this was done it was found that there was a defimte negative temperature coefficient, i.e, the rate of reaction pro- ducing the glow, mcreases as the tempera- ture is lowered For the particular e:xperiment quoted, Rayleigh found the reaction rate to vary as T-o·s• over the extreme range of temperature investi- gated.

8. SPECTRUM OF THE AFTERGLOW

M.oleeular mtrogen has the followmg band systems m the VISible and the near mfra red and ultrav1olet reg10ns (see F1g. 18).

1. First pos1hve group A~B, ;\;\ 10491-5371.

2. Seeond pos1hve group B~C, ;\;\ 2814-4917.

3. Fourth pos1hve group B~D. ;\;\ 2900-2JOO

4 F1fth poslt1ve group (Gaydon, 1944a) ;\;\ 2781-2033.

5 Vegard Kaplan mter combi-

nation system (forbidden X~A. ;\;\ 3321-2332 t :fu.ns1 t10n)

6. Fust negat1ve bands (due to N₂+) x·~A', ;\,\ 5864--2987.

(Note-Pos1hve bands are observed strongly m the pos1bve column and the negative bands m the negat1ve glow near the cathode of the d1scharge tube )

The band~ (1), (2), (3), (4) and (6) are excited to different degrees of intensity according to the strength of the exciting diseharge. (5) is ob~ervc-1

in the emission from nig~ht sky. It can also be obtained in the laboratory, but <mly under special condition of excitation all61 with specially prepared t11scharge tube.

The spectrum of the afterglow is to be carefully distinguished from that of the glow in the disehar~e. In the latter any or all of the bands listed above may be developed, and, unless prel.'autions are taken, they may contaminate the spectrum of the pure after- glow. It is not always enough to observe the afterglow speetrum at

~me distance from the actual discharge, because the exciting field may extend well beyond the region in which the glow is observed.

The safest method of ensuring that the speetrum of the pure after- glow is being observed is to stop the discharge. When so obsrrved the spectrum of the afterglow is found to consist of selected bands of the First positive group only. They are in the red, the green and the yellow and give the afterglow its characteristic golden yellow colour. The wave lengths of the prominent bands are:

red 6251

yellow 5802

green 5371

In the table below the wave lengths of the bands of the First positive group, with their vibration levels are given. The bands enhanced in the afterglow are in bold figures.

v" I 1

~ ⁰

11213

^{4 5 6 / 7 t. 9}

-i-=:~~::~57

2 - 7753'05 8722'28

·_ J

¹9927 ^II

=- -- I ^{--! ~ ,_---}

I

3- =

7626'131s541'73 19657

~ ==- ⁼⁼⁼ ==I= --

4--

---1~

⁷⁵⁰⁴ 8369'0! 9409 - - - - ____

! ___ - - -

_5_ --~~~7386'5 ~204'62 918~- _· _ _ , _ _ ._i_

--=---

_: ^==~~

⁶⁰⁷⁰

^--,:~~:

^::::

^~~~:~:

^{6 7896'28}

^{1 - -=--=- =} ==

8 \ I

I

⁵⁹⁵⁹

16469

-9 -=-=~==~==-- ~ ^{5906 6394}

_-_--1_-_ -- ---

~~---\---1---i=---== __

^{5442 5855 6S2S ---}

~ ~ ^{_ 1 I 1}1H07 5804 16252 - -

12 I --~---,---~--- ---~---~s:J75- ~755 6185•

It should be mentioned that systematic search in the •infra-red region of the afterglow spectrum has not shown the existence of any band systems otb':!r than those of the First positive group ( Kichlu and Acharya, 19Z8). Searc:1 in the ,·acuum ultra violet also showed that there is no band system which could be attributed to the afterglow (Sponer, 1927; Johnson and Jenkins, 1926).

The intensity dstribution in the bands of the afterglow spectrum is greatly modified by the admixture of other gases, b~· change of pressure and also by temperature. This is illustrated in Fig. 11.

I I

f

I

~

~,..,.,

I.

^I

_' t

r

t

::>!lt~2

:\--"':"'"'7'~

IIIII ^{I I}

¹

^f

^J

Ill ' ^{Ill (a>}

Ill,,, ^{dlt 1111}

1111 ..

¹

lfllllse

^(CJ

II,'

^{I (d)}

Fig. 11. Illustrating ho" the intensity distribution in the Ftrst positive group (Nz) varies under varied experimental conditions. (a) Positive column in ordJI!lary discharge. (b) Afterglow in 100% ~outrogen. (c) Afterglow with 75% helium and 25% mtrogen. (d) Afterglow at the temperature of liquid rur. (a), (b) and (c) are after Rayleigh. (d) is after Herzberg.

(For wavelengths of the hnes see Table in page 23.)

In the figure, (a) is the spectrum of ordinar~· glow diseharge in nitrogen in a capillary tube showing the First positiw bands. The enhancement. of selected bands in the afterglow spectrum is shown in (b). In (c) is shown how the intensity distribution in the after- glow is modified when helium is present with nitrogen (Rayleigh.

1922). The effect of low temperature on the intensity distribution is shown in (d) (Herzberg, 1928 b).

9. EXOITATION OF SPECTRA

Many experiments have been made on the excitation of the spectra of gasses and \'apours intr·oduced into the glowing gas

Active nitrogen generated in a discharge tube is continuously pumped out through an observation tube and the vapour of the substance to be studied is introduced a little above the la\ter.

Generally, the luminescence •due to tjle vapour of the gas replaces the golden yellow afterglow. The observation tube is viewed end on through the spectroscope. For the study of solids (metals) which cannot be vaporised, a method due to Rayleigh (1916) can be employed. The metal is used as a cathode when it is "scattered" and the "scattered" metal in the stream of the glowing gas emits the characteristic spectrum of the metal.

The spectra excited by the glowing gas differs in many res- pects from those developed by the electric arc. This is be- cause, as in the case of the First positive system, of N2 , the intensity distribution 'm the lines or bands is radically modified. Generally, in each band fewer lines are developed. Further, neighbouring bands are not overlapped to the same degree, so that the heads stand out more prominently than in the arc spectrum (Jevons, 1932).

Excitation by active nitrogen is thus" greatly advantageous in the study of spectrum.

The spectra excited are those of,. (a) the substances introduced, (b) the compounds (nitrides) formed by chemical action and (c) the dissociated products.

(a) In the first category we have the spectra of metallic vapours. These are intermediate in character between the arc and the flame spectra (Jevons, 1932). Na, K, Cd, Zn, Mg and Th have been studied by Okubo and Hamada (1928). (Attempts by Okubo and Hamada to obtain spectra of Ca failed.) AI, Pt, Cu, W and Ni have been studied by ·'scattering" in uncondensed discharge by Rayleigh (1916).

Amongst the gases studied CO and NO are ex<'ited to yield the

<:>haracteristic spectra (Knauss, 1928). For CO a few bands of the Fourth positive group are obtained. These correspona to vibration levels 1-6 of the A-state. (The energy is 8'2 to 9'0 e V). The eyanogen bands are also developed. For NO the {3, y and

o

bands

<lre developed.

H2 and 02 are not excited by active nitrogen.

Amonbst eompounds 12 , SnCI. and Hgl yield spe~tra re- sembling those obtained in vacuum tubes (Strutt and Fowler, 1912).

The more refrangible parts of the spectrum are, however, developed to a greater degree in the afterglow.

Ti('l4 ~ives its arc spectrum (.Tevons, 1914). Spectrum of AuC'l. has been studied by Ferguson ( 1928) .

. Speetra of eopp\'!r halides (CuCl, CuBr, Cui) excited by aetive nitrogen, have beep studied ~y Mulliken (1924).

(b) In the seeond category, excitation of the compounds formed (nitrides formed by chemieal action )-the most common example is that of nitric oxide. Oxygen present in the minutest trace yields the spectra not of 0 2 but of NO ^({3 and y bands, Knauss, 1928). The spectrum of NO bands, excited by acth·e nitrogen in discharge in helium with one per cent nitrogen, and oxygen as the slightest trace of impurity has been studied by Anand (1943).

Effect of introducing NO in small quantity in the afterglow is the same as that of introducing air, i.e., f3 and y bands are deve- loped (Rayligh, 1916-17). If NO be introduced in large quantity a continuous band is produced in the less refrangible part of the spectrum This might be compared with the continuous spectrum of nitric oxide in the blowpipe flame. It might be mentioned in this connection that though NO hands are developed easily even when 02 is present in very small quantity, the outflowing gas may not yield nitric oxide in detectable amount by chemical reaction. In fact, it was for this reason that the' identity and the origin of f3 bands-- whether it was emitted by active nitrogen or by a compound of nitrogen and oxygen was for a long time uncertain.

SiCl4 (,Jevons, 1914) and BC13 (Jevons, 1915) yield the spectra of the corresponding nitrides.

(c) Amongst the spectra developed of dissociated products mention may be made of HI and HBr. These introduced into the glowing gas ~-ield spectra of 12 and Br2 respectively. (HCl does not ~-ield any spectra.) Hydrocarbons in general produce strongly the CN bands, though the chemical yield is hydrocyanic aeid.

10. CHEMICAL ACTIVITY

Rayleigh ( 1911) was the first to draw attention to the chemi- cally actin• nature of the glowing gas and the name "Active Nitro- gen'' is due to him. The chemical action on solids like phosphoru,s can be studied by drawing the glowing gas over the1 substance. For studying the action on gases, the gas to be examined may be led into a stream of glowing nitrogen and the resultant product collected by cooling by liquid air and examined. For studying the effect on