;S6 



NA TURK 



[July iS, 1S95 



(UliVrent from its properties in the primitive state. On reloading 

 the overstrained rod it is found that the proportionaHty of strain 

 to stress no longer holds good, even under verj- light loads, and 

 further that there is " creeping," or continued extension with the 

 lapse of time, when any load is kept on for a few minutes. 

 .\gain, on removing load the bar continues to retract for some 

 time. These features of the overstrained state are most con- 

 spicuous in tests made directly after the overstrain has taken 

 place. They tend to disappear if the bar is allowed to rest for 

 some days or weeks. This elastic recover)' with the lapse of 

 time, some features of which have l>een already noted by 

 Kauschinger and others, is less rapid in moderately hard steel 

 than in iron or mild steel, ap|>arently because the condition of 

 iwerstrain requires a greater load to produce it. Thus a rod of 

 common iron, overstrained so much that the yield-point was 

 reached, was found to have made a practically complete recover)- 

 of its elasticity in five days. On the other hand, in a rod of 

 rather hard steel, overstrained by appl)-ing a load of 1 1 tons and 

 >ul)sequently tested with loads of S tons only, the recover)' was 

 still impcrlect after three weeks. The following table shows 

 the progress of the recovery by giving the observed extensions of 

 this rod after three intervals, namely ten minutes, one day, and 

 lliree weeks, after the overstrain took place. 



The molecular settlement which is shown by these experiments 

 to lie going on for some time after overstrain has taken place, is 

 known to be associated with a rise in the yield-point. Instances 

 of this were given by the author in a prev-ious paper {Proi, Roy. 

 .S'(V., No. 205, 1880). 



May 30 — " On the Motions of and within Molecules : and 

 • in the Significance of the Ratio of the Two -Specific Meats in 

 (iaseJv" By Dr. G. Johnstone .Stoney, K. K^S. 



In treating of molecular physics it is found to lie convenient to 

 widen the meaning of the word motion, so thai it may lie 

 employed in regard to any change or event in which encrg)' is 

 -torcd, whether as kinetic cnerg)', or as potential, electrical, 

 chemical, or any other. It is in this generalised .sense that the 

 term is lo l>e understo<j<l throughout this pa|>er. 



The aim of the ia|>er is lo demonstrate the existence of events 

 going on within the molecules of matter which are so sluggish 

 in affecting its pressure when in the gaseous state, or its tem- 

 l^eralure as measured by the thermometer, th.-it il is only after 

 millions of encounters that any manifestation of their having thus 

 lost cnerg)' by conduction t>ec<imes appreciable ; while these same 

 events are prompt and active agents in other o|K'rations of nature 

 'hrough chemical reactions or l>y radiation. 



Molecular events may lie <listinguishe<I into A or external 

 events, and B or internal. The external events arc the move- 

 iiii'nls of the centres of inertia of the molecules relatively to one 

 .itii.ther. They present themselves most conspicuously in those 

 1 •mpar.ilively protracted journeys which the molecules of gases 

 m^\,. |,i„,.,.n (heir much briefer encounters. By B motions are 

 ' ' :dl events in which cnerg)' can be stored ihal 



lividual molecules, including rotation of the 

 ' nny movement of this kind, which, how- 

 long wilh every other relative motion of 

 ^uule: movements within its |i<in<lerable 

 matter, or of its electrons, changes in the configuration of its 

 parts, and ever)- other event within the molecule which can 

 al'S'irb and yiel'f encrg)'. The electrons arc those remarkable 

 ch.irges of electricity, all of the same amount, which are asso- 



NO. 1,342, VOL. 52] 



ciatetl in ever)' chemical atom with each capiicity that it possesses 

 of entering into combination with either atoms. 



It is convenient to distinguish the B or internal events, into 

 Ba events between which and the .-V or translational motions of 

 the molecules there is ready interchange of energ)' whenever 

 encounters take place ; Be events which are so isolated that no 

 such interchange takes place ; and Bb events which lie between 

 these extremes. In the struggle which takes ])lace during an 

 encounter, or in arf)' one of the much longer intervals between 

 two encounters, a Kb event will part wilh but ver)' little of any 

 excess of enei^' it may possess by conduction, /.('. by trans- 

 ferring energ)' over to .\ or Ba events. Nevertheless it may 

 sustain an appreciable loss of energy in this way when the mole- 

 cule has been buffeted in a sullicicnt number of encounters. This 

 may easily occtir in a time which seems short to us, since, if th( 

 gas l>e at atmospheric tem]>erature and pressure, each molecule 

 meets with some thousands of millions of encounters ever) 

 second. Meanwhile, during this process, which is slow from 

 the molecular standpoint, the Bb events, if they have electrons 

 associated wilh them, may be engaged in a prompt and active 

 exchange of energ)' with the a.lher by radiation. 



In substances that are appreciably phosphorescent, it is easy 

 to detect the presence of these Bb events ; and, accordingly, a 

 proof that they exist in this class of bodies is given in the jxaper. 

 Moreover, by comjaring the behaviour of different phos- 

 phorescent bodies, we learn that the degree of isolation in which 

 Bb motions stand varies much from substance to substance. 

 Motions of this tyi^e, which are so conspicuous in the bodies 

 that can be perceived to be phosphorescent, are, of course, not 

 confined to that class of bodies. In fact, they appear to be an 

 important part of what is going on in every molecule of matter 

 that can emit a S|)ectrum, a description which probably em- 

 braces ever)' molecule. 



Since Bb motions are in various degrees isolated from the 

 other events that are simultaneously going on in the molecules, 

 it follows that in some gases the specific heat as determined by 

 experiment will not be a definite quantity, but will [xtrtly depend 

 on the duration of the experiment by which it is determined — 

 i.e. ujwn whether or not there has been time for an interchange 

 of energy Ijetween the Bb motions and the .-^ and Ba events. 

 This is likely in some gases to make an appreciable difference 

 between determiniitions of 7 — the ratio of the two specific heats 

 — deduced from the observed velocity of sound in the gas (where 

 the real experiment lasts only during one semi-vibnition of the 

 musical note employed), and determinations made by other 

 experiments which require seconds, perha]>s minutes, to carry 

 them through. 



There is rea-son to believe that it is with these Bb motions that 

 the electrons within chemical atoms are chiefly a.ss()ciated. and 

 that in most c;ises il is ihey which are concernetl in luinintjus 

 effects, whether in flames or when the gas is under the influence 

 of electricity. Accordingly in both cases the luminous etTects 

 may have their origin in events that are in a consideral)le degree 

 isolateil from those that directly affect the ihermiuneter ; and 

 wherever this is the case, the luminous eft'ecls will be in excess of 

 what belongs to the temperature of the gas as determined by its 

 power of communicating heat by conduction to bmlies upon 

 which ils molecules impinge. This seems to have been proved 

 l>y I'rof. Lewes of flames {Pniiiidings of the koyal Society, 

 vol. Ivii. p. 404 and p. 467), and many phenomena indicate thai 

 it is also true of all ga.ses which exhibit spectra of bright lines 

 when in that slate which has been mi.scalled incandescent. 



Il is s|K"cially lo lie noted that the interpretation usually put 

 upon the value of y in a gas has to be profoundly modified in con- 

 sequence of the presence of Bb motions within the molecules, 

 and of the degree in which the corresponding Bb motions of 

 swarms of molecules are more or less linked together by the 

 interaction that goes on between their associated electrons and 

 the .-vlhcr. (Sec I'"il?.gerald, in the Prodcdings of the Koyal 

 Scx;icty, vol, Ivii. p. 312.) 



These examples may serve to show how a knowledge of the 

 presence and .activity of Bb motions supplies a clue lo interprct- 

 mg some of the jihenomena of nature ; and the extent of its 

 applications may \k. judged by reflecting that it is electrons 

 for the most \y\\\. .associated with Bb motions which apjK'ar to 

 be printarily concerned in every chemical reaction and in all 

 ]>hcnomena of radiation. 



"On the \'elocities of the Ions." By \V. C Dampier 

 Whctham. A continuation of a former i>a|x;r (Phil. Tram. 

 184, 1893 ^< P- 337)- The velocities of certain ions 



