July 26, 1883] 



NA TURE 



297 



However many times material atoms may be hyper-attenuated 

 or condensed, their substance no doubt retains its original 

 material status^ although removed by numbers of grades or 

 orders of attenuation from it to which the mathematicil prin- 

 ciples of the theory assign no limit ; and boundless space is thus 

 strewn at once with a grade of common-matter atoms, which in 

 their original status may have properly belonged to any other 

 grade of unknown remoteness. But this fixity of matter's original 

 grades of size and density with only infinite insulations from other 

 grades, is not more notable than the unrotativeness or fixed direc- 

 tions of some coordinate axes of mechanical motions in space 

 which does not prevent the motions from being just as perfectly 

 describable by the selection of any other equally fixed ones. We 

 are in the same way unable to say by how many revolutions the 

 hands of a clock have reached a certain position on its dial, 

 unless we examine and properly employ to estimate it the state 

 of wear and attrition of the wheel-work of the clock's driving 

 train, or unless we know the number of times that the clock had 

 been wound up. 



The solution of some very bewildering physical questions is 

 offered by this hypothesis ' when we reflect, as I have before 

 endeavoured to explain, that the expansions here considered are 

 all of them variations of a quantity <f> (or "entropy" of a 

 homogeneous body at the same temperature throughout), which, 

 by its mathematical description, is obviously the rati 3 index of a 

 describing point's place upon a hyperbola, and which therefore 

 passes continuously through an endless series of values o and --o 

 (which revisit each other in graphic space, just as a circle-radius 

 revisits its former place after every passage through four suc- 

 cessive right angles), while the describing point pursues the curve 

 continuously. 



There is enough evidence in geometry to show that this hyper- 

 bolic variable of position, and the angular one on the hyper- 

 bola's auxiliary circle of a certain configurated point on that 

 circle, cannot pursue their geometrically configurated course 

 together through more than a quadrant of the circle and hyper- 

 bola from the two curves' common apex without violating the 

 axioms of ordinary geometry. Thus it is clear that in the 

 transition state of the measure <p through infinity from one 

 "grade" of a mass's state of attenuation to another, there is 

 needed a new law of geometry (or at least of continuous material 

 motion) allowing a new pair of tracing-points supplanting the 

 disused former pair at each dead-point of the two curve-, to 

 describe a new quadrant of the hyperbola and of its auxiliary 

 circle from that point, with a constant geometrical configuration 

 to each other without violating geo netrical axioms. 



This transition law and the nature of the c tnfiguration which 

 it frees from geometrical contradictions while giving it con- 

 tinuous validity round the whole circuit of the circle and hyper- 

 bola together, is so exactly what has just been de-cribed of the 

 nature of material points' or of physical integrant-parts' compo- 

 siteness while still remaining points in their m >t ir properties, 

 that almost all reason for doubt and question seems to be ex- 

 cluded that it is the sought-for law and mode of motor connec- 

 tion between 9 and <f> (or angle- and entropy-position of a point 

 or homogeneous body), which links universal heat-motion of 

 natter to all those other, no doubt therefrom derivable lout 

 otherwise unaccountable descriptions of matter's motions which 

 we see in physics. 



On Lord Rayleigh's Dark Plane 



In Nature, vol. xxviii. p. 139, was printed a communication 

 from Lord Rayleigh to the Royal Society on the subject of the 

 dark plane which is seen above hot bodies in dusty and illumin- 

 ated air, and which had long been used by Tyndall, and after 

 him by science te chers generally, as an illustration of the fact 

 that light which does not enter the eye cannot be ^een. 



It had never occurred to me to doubt the validity of the com- 

 uo onlv-received explanation of the dust-free space, viz. that the 

 dust in the dark region had been either burnt up or dried up by 

 contact with the hot body, and I was struck and greatly inter- 

 ested in the definite character of the phenomenon as described 

 by Lord Rayleigh in your pages, and in his conclusive shattering 

 of the old explanation by the simple device of using a cold body 



1 In particular, as will be easily gathered from the above brief comments, 



of the law of dissipation or of a fixed tendency to gradual reduction and to 



rsaj uniform diffusion of all forms of energy in a given link of matter's 



in one common form of the energy of heat, or of the worjo; of entropy- 



expinsion. 



instead of a hot, and so getting a down-streaming dust-free 

 space instead of an up-streaming. 



I ns however quite unable to accept Lord Rayleigh's very 

 tentative hypothesis that the curvature of the stream-lines and 

 consequent centrifugal actions might possibly account for the 

 phenomenon, nor do I imagine that he himself ever regarded 

 this as anything more than a guess thrown out for want of a 

 better. 



I mentioned the matter to Mr. J. W. Clark, whose services as 

 Demonstrator I have lately had the good fortune to secure, and 

 he proceeded to make a few simple experiments with a view 

 first of repeating the observation, and next of testing an elec- 

 trical hypothesis which suggested itself. 



The hypothesis is one that has failed to verify itself, but it 

 may be just worth stating. The difference of temperature between 

 the solid and the air causes convection currents, the air thus made 

 to stream over the surface of the solid electrifies itself by friction, 

 and the dust particles are expelled from the electrified air. 



We were early led to dou'it whether the insignificant amount 

 of friction which alone was acting in some cases could possibly 

 produce the effect ; an 1 in fact it was soon found that though 

 electrific tion modified the phenomenon it pretty certainly did not 

 cause it. 



A doubt then arose whether the space was actually dust free 

 or only optically so ; whether anything like mirage Mue to 

 unequal densities could account for the darkness. These ideas, 

 however, would not hear consideration, and we soon convinced 

 ourselves that the region is really transparent air free from dust, 

 though its extreme sharpness and blackness render it difficult at 

 times to refrain from thinking of it as a black opaque film. 



Irregular dark stria: obviously allied to the regular dark plane 

 are to be perpetually observed in any duty air disturbed by con- 

 vection currents ; and nothing but the want of the necessary 

 illuminati 111 prevents our commonly ob-erving what must be one 

 of the mo^t universal appearances, viz. dust-free regions stream- 

 ing from every solid bo 



We are n iw pretty well convinced that differences of tempera- 

 ture have nothing to &o with the real nature of the phenomeno 1 ; 

 we find thai solid bodus have sharply-defined dust-free coats or 

 film! of uniform thickness always surrounding them, and that 

 these coats can be continually taken off them, and as continually 

 renewed, by any current of air. The slightest elevation of tem- 

 perature of the solid causes its dark coat to stream upwards ; 

 the slightest depression of temperature below that of the atmo- 

 sphere causes the coat to stream downwards ; but the coat is 

 there all the time, independent of convection curren's, though I 

 believe it gets thicker as the body gets warmer. Why the air 

 near a solid is free from dust we are not prepared to say. 



A few of our earlier experiments might readily enough have 

 suggested the old exploled explanation that the smoke was 

 either burnt up or dried up or otherwise temporarily rendered 

 invi-ible by heat. Take for instance a long piece of ordinary 

 quill glass tubing ; blow it half full of roloacco-smoke, and 

 hold it horizontally in a beam of light. The first thing to 

 notice is the curious way the end of the stream of smoke draws 

 out to a point with a sharply defined edge, and how it falls 

 about in-ide the tu'>ewhen the tube is rotated. Next warm a 

 part of the tube gently : a space clear of smoke at once appears 

 and widens. Next heat the tube in the flame of a Bunsen and 

 blow smoke gently and continually through it : the smoke narrows 

 down to a mere thread as it passes the hot place, or it may disap- 

 pear altogether in a pointed cone ; but it reappears on the other 

 side of the hot place, and it issues from the end of the tube. 



Our experiments have been mostly conducted in a glazed 

 cigar-box with one or more horizontal copper rods passing into 

 it through insulating glass tubes, the ends of the rods carrying 

 binding- screws into which could be clamped scraps of sheet 

 copper of various shapes. The illumination was either sunlight 

 or an oxyhydrogen lamp, or more usually, and far the most con- 

 veniently, a Serrin arc-lamp in its lantern, fed by a secondary 

 The smoke employed was nearly always tobacco, for 

 we soon satisfied ourselves that the nature of the smoke or dust 

 did not affect the essence of the phenomenon, and we conse- 

 quently used that which was the easiest and for which the im- 

 plements were always at hand. Sal-ammoniac was, however, 

 occasionally used instead. 



It was wholly unnecessary to heat the rod in order to start the 

 dark up-current, for if it is not infinitesimally warmer than the 

 air to begin with, the beam of light will warm it sufficiently in 

 an it stant. Still the rods can be healed by a lamp outside the 



