536 



NATURE 



{April a,, 1889 



be one of degree rather than of principle. For, even if its appli- 

 cation for the full time required for that stage of evolution, in 

 the case of so large a nebula as the original solar one, were 

 considered doubtful, there is no objection apparently on that 

 head to be adduced in the case of smaller nebulas, or comets (the 

 smallest class of nebulae) ; in which latter case there is some 

 support to the theory afforded by experiment and observation. 



In regard to the question of " elasticity," the only resource, 

 in my view, is to abandon this idea, in the ordinary sense of that 

 term (which conveys the idea of retention of form), and suppose 

 that there may be complete disintegration by the collisions at 

 times, ' welding or fusion together at others ; so that the mean 

 degree of aggregation remains constant so long as the trans- 

 latory motion remains constant. This idea of the existence of 

 a number of possible mean states of aggregation of matter 

 between the extremes of complete integration into one mass, 

 and complete disintegration into molecules (the states being de- 

 pendent on the rate of translatory motion), was thrown out 

 by me in Nature (vol. xix. p. 461);* also, further, in the 

 Philosophical Magazine, August 1879, p. 153. 



According to this view, it is implied that " cohesion," as a 

 central force, can play the same part as "chemical action" 

 under translatory motion, and produce fluctuations about a 

 mean state of aggregation— just as in a compound gas, for ex- 

 ample, even at normal temperature, the small lumps of matter 

 which move as wholes in the motion of translation arc dis- 

 integrated now and then, and integrated elsewhere, the mean 

 state (only) of the gas remaining unchanged. 



The above, with the exception of the last two paragraphs, 

 was written before the appearance of Prof. G. H. Darwin's 

 reply (March 14, p. 460), to some criticism I ventured to offer 

 (March 7, p. 436) on certain points of his theory in my first 

 letter. I would make a few remarks in addition here. 



I am not quite able to agree with the view of M. E. Minary, 

 expressed in a paper brought before the French Academy on 

 February 18, of which an abstract appeared in Nature of 

 February 28, p. 432, and is referred to in Prof. Darwin's letter. 

 The chief passage, as given in the abstract, is as follows : — 



" The gases being perfectly elastic bodies, and in the upper 

 atmospheric regions in an extremely rarefied state, heat cannot 

 be produced by the shock of bodies endowed with great velocity 

 and impinging on perfectly elastic molecules capable of receiving 

 the motion and acquiring the velocity of these bodies : in 

 this case the movement is communicated, not dissipated or 

 transformed into heat." 



This view is opposed, I venture to think, to the deduction we 

 may draw from such an experiment as that with the "fire 

 syringe," where the air in a cylinder is infiamed by suddenly 

 compressing a piston. What is the flame here observed through 

 the tube of glass due to ? It is due, of course, to the vibrations of 

 the molecules of air, which break the ether up into waves, and 

 so affect the eye. From the fact that the air molecules are 

 "perfectly elastic," it becomes impossible for a moving body to 

 impinge against them violently without throwing the molecules 

 into energetic vibration, which is the physical basis of "radiant 

 heat." The same must occur on a greater scale, as it appears, 

 when the air is compressed by a flying meteorite, although I 

 accept M. Cornu's suggestion (quoted in the abstract, p. 432) 

 that the luminosity observed may partly be of electric origin. 



In Joule's "Scientific Papers," vol. ii., experiments in 

 association with Sir William Thomson are described, of whirling 

 thermometers through the air (attached to a lathe). The 

 experiments, which were numerous, gave 1637 as the velocity 

 in feet per second ; on the average, equivalent to a rise of tem- 

 perature of 1° C. (vol. ii. p. 316). Whirling a thermo-electric 

 junction attached to a reflecting galvanometer was tried with 

 consistently the same result as the thermometer (p. 310). These 

 experiments were made partly * with the view to test the theory 



' A special property of iron, which may have importance here. 



^ It would seem even curious to my mind, if there were no intermediate 

 state between these two extremes of complete integration and complete dis- 

 integration, by a varying rate of translatory motion (or energy). It may be 

 observed that the radiant energy set free so abundantly at a collision is not 

 lost, but radiated in great part to another region in the nebula (and there 

 absorbed). 



3 Also see vol. i. pp. 399, 536. The temperature w.is found to be inde- 

 pendent cf the form and size of the thermo-electric function, and was assumed 

 as evidently independent of the density of the air. No doubt the more air 

 there is the more there is to heat. But it nevertheless seems plain, I think, 

 that the temperature of a meteorite must rise higher in dense air than in 

 excessively rarefied air. How is this to be explained ? ■ Each individual mole- 

 cule of air in striking the moving meteorite, is thrown into violent vibration, and 

 this (temperature) is independent of the number of air molecules evidently. 



of the heating of meteorites. The temperature was fo\md to be 

 as the square of the velocity. The law of Clausius, that the 

 translatory motion and the vibratory motion (which latter motion 

 alone affects the eye and senses as radiant heat) of the molecules 

 of an ordinary gas are proportional to each other, has — as Prof. 

 Darwin allows — been experimentally verified through^ a con- 

 siderable range of temperature. To my mind it appears obvious 

 that these two forms of motion (translatory and vibratory) must 

 be interconvertible and mutually sustain each other. When the 

 gas, for example, is exposed to the pulsations of ether waves 

 (radiant heat), this vibratory motion is first taken up by the 

 molecules, but part of it is converted into translatory motion, 

 as proved by the rise of pressure. If, on the other hand, the 

 translatory motion of the molecules of gas be augmented, part 

 of this is instantly, as we know, converted into vibratory 

 motion, the source of radiant heat. S. ToLVER Preston. 



Paris, March. 



The Molecular Formulae of Aluminium Compounds. 



In a letter to Nature, December 27, 1888 (p. 198), I gave 

 a tabulated statement of the numerous vapour-density determina- 

 tions of halogen, and a few other compounds of aluminium and 

 the allied metals, and pointed out what appear to me to be the 

 legitimate conclusions to be derived from the experiments, 

 regarding the molecular formulae of these compounds. 



Since then two interesting articles have appeared in Nature 

 (pp. 447 and 495), in which accounts are given of determina- 

 tions of the vapour-densities of aluminium acetyl acetonate, 

 A1(C5H70.2)3, and aluminium methide, [A1(CH3)3]„. The simple 

 formula given for the first compound has been proved to be 

 correct, at any rate for the conditions under which the ex- 

 periments were made. The results given for aluminium methide 

 are : calculated for Al2(CH3)6, 4*98 ; calculated for Al(CIl3)3, 

 249; observed (10° above the boiling-point under atmospheric 

 pressure), 3*92. 



What I ask permission to call attention to and to criticize is 

 the conclusion drawn by the author of these articles from the 

 experimental results. Speaking generally, the conclusion may 

 be stated in this way : Molecules of the formula MR3 do exist, 

 therefore molecules with the double formula MgR,, do not. I 

 confess that I am wholly unable to appreciate the force of the 

 argument. Must we take the existence of the molecule NOj as 

 a proof of the non-existence of N.2O4? 



That I have not stated the argument unfairly may be shown by 

 quotations from the article (p. 495). Speaking of aluminium 

 acetyl acetonate the author says: "It is supremely satisfactory 

 that in this case the density, at a temperature only 45° above 

 the boiling-point, was found to actually correspond precisely 

 with that required by the triad formula, precluding again the 

 possibility of the existence of molecules of the type AlgRg-" 

 And previously, after giving the results for aluminium methide, 

 notwithstanding the fact that the observed density 3 '92, obtained 

 by Quincke, corresponds rather more closely to the higher than 

 the lower formula, the author remarks : ^^ Hence it can no longer 

 be doubted that molecules of the double formula are incapable of 

 existence." The italics in both cases are mine. The conclusion 

 I should draw from all the experiments with aluminium com- 

 pounds is this : The experiments of Deville and Troost, Friedel 

 and Crafts, and Louise and Roux, prove conclusively that 

 molecules of the higher formula AlgRg are capable of existence ; 

 the results obtained by Nilson and Pettersson, and by Buckton 

 and Odling, point also to the existence of molecules of the lower 

 formula, but further proof was certainly needed, and this has 

 now been afforded by the valuable experiments of M. Alphonse 

 Combes with aluminium acetyl acetonate. 



But in dense air. no doubt, the heat accumulates much faster than it can be 

 radiated away, and so the temperature of the meteorite attains a final maxi- 

 mum, which is greater the denser the air is. The temperature may probably 

 be appreciably constant within certain limits of variation of density ; but it 

 appears obvious that in excessively rare air the temperature developed must 

 be less, owing to rapid dissipation in space by radiation. 



' The ratiO of Maxwell referred to in my last letter, and which Prof. 

 Darwin remarks in a footnote (p. 460) was stated inaccurately by me — was 

 not, I may explain, the one quoted by him from Maxwell's " Theory of 

 Heat." It is contained in Maxwell's paper "On the Dynamical Evidence 

 of the Molecular Constitution of Bodies " (Nature, vol. xi. pp. 357 and 374). 

 After alluding to the l.aw of Clausius, Maxwell remarks: — "'In i860 I 

 investigated the ratio of the two parts of the energy on the hypothesis that 

 the molecules are elastic bodies of invariable form. I found, to my great 

 surprise, that, whatever the shape of the molecules, provided they are not 

 perfectly smooth and spherical, the tatio of the two parts of the energy must 

 be always the same, the two parts being in fact equal" (p. 375). 



