222 



NATURE 



[January 3, 1895 



be considered to be ia simultaneous collision with one another. 

 Thi^maybe where the spectrum crux fails. Perhaps somebody 

 would be so very kind as to point out where exactly in these 

 generalised coordinate investigations the postulate of collisions in 

 pairs is used, and so save lazy people like me the trouble of 

 hunting it up. 



This raises the third point as to how this difficulty abaut 

 the spectral lines is to be surmounted. I cannot follow either 

 Mr. Bryan's, or what I understood to be Mr. Larmor's view, 

 tha*. any help can be got by supposing spectral lines to be due 

 to electromagnetic vibrations. The example Mr. Bryan gives 

 of smooth solids of revolution is quite beside the point. In 

 this case there is no interchange of energy between rotation 

 round the axis of revolution and the other degrees of freedom. 

 This is quite contrary to what we know to be the case in respect 

 of ethereal and molecular energy. We know that radiations 

 cause bodies to become cooler, and therefore there is inter- 

 change of energy. This could not be otherwise, as is evident 

 from what we know of the mcch.inical forces — electric, magnetic, 

 ami electromagnetic — that interact between matter and ether. 

 It is rather hard for Mr. Bryan to say that the onus probandi 

 lies with f hysicisls to explain exactly hou' transference takes 

 place ; surely the fact that transference does take place is 

 sufficient to prove that a complete theory should take in both sets 

 of coordinates — ethereal as well as material — and I should have 

 thought that those formidable arrays of <ip^ dp, . . . dp,, &c., 

 of coordinate; and momenta to the «th, with dots between to 

 signify their indefinite number, should include everything of 

 this kind that could possibly be required. Here, however, the 

 postulate of collisions in pairs entirely breaks down, and thus 

 shows a way out of this spectral difficulty. A second way was 

 suggested to me, by whom I forget, at Oxford, namely, that the 

 complicated systems of lines that we see in the spectrum — of 

 iron, (or instance — are so connected as to their amplitudes, 

 periods, and phases, as to represent only a single coordinate. 

 Anybody who has tried to expand a simple function in Fourier 

 series will easily understand how a very simple motion might 

 produce a fearfully complicated spectrum. 



It seems, then, to me that the questions which need solving in 

 our study of the dynamical foundations of thermodynamics at 

 present are (i) how to explain spectra, and (2) how to deal with 

 several bodies in simultaneous collision ? 



Geo. Fras. Fitzgerald. 



Trinity College, Dublin, December 19, 1894. 



The difficulties in the way of harmonising the spectro- 

 scopic phenomena of healed gases with the conclusions of the 

 Boltzmann-Maxwell theorem of distribution of kinetic energy, 

 appear to mc to have been fxaggerated. 



This theorem asserts, as a purely dynamical proposition, that 

 a very large number of, say, billiard balls, perfect as to spheri- 

 city and elasticity, having been set in haphazard motion in a 

 space bounded or interrupted by perfectly elastic rigid surfaces, 

 if the volume of free space be sufficiently large as compared 

 with the total volume of the balls ' will, if left to itself, lend 

 to a certain stale of density and velocity distribution, and 

 having reached this stale, will remain in it permanently. The 

 billiard balls may be replaced by rigid bodies of any form, or 

 indeed by material systems capable of any changes of shape or 

 motion of the parts among themselves, and a corresponding 

 proposition still prevails — one general property holding good in 

 all cases, viz. that if u, v, w he the motions of translation of 

 each system as a whole, m its mass, T its total kinetic energy, 

 and » the total number of its degrees of freedom inclusive of the 

 3 of translation, then in the permanent state 



2 



mv- 



2 



J = T/«, 



where - denotes mean value throughout the whole medium. 



/'/iOT'//aciV there is a discontinuity in thij abstract proposi- 

 tion anfi ting it for the basis of physical investigation, e./^. the 

 imalleit possible want of perfect sphericity in the billard 

 ballt would appear to effect as complete a change in the 

 physical properties of the medium as if each ball became an 



ellip-oid or tetrahedron, the mean being changed /fr 



' of ca4C!i in which three or more 

 infinitely smaller (ban the number 



solium from T/3 to T 6> and the difficulties introduced by the 

 spectroscope are founded upon this discontinui y. Doubtless 

 we cannot make n fractional, but it should be remembered that 

 the dynamical proposition only speaks of an ultimate state, and 

 ignores the rale of approach to that state. The continuity of 

 physical properties is maintained by attending to the continuity 

 of change in this rate. This rate I have estimated for a 

 particular case in the new edition of my short treatise on the 

 subject, showing that there may be a sensible permanence long 

 before the law of equal partition is established. 



And, again, there is yet another point to be considered. The 

 passage from the thermal to the optical properties resembles 

 the passage from mere noise to music. Dynamically it is the 

 passage from irregular, haphazard motion, heat, to regular 

 periodic motion, light ; the former must be decomposed into its 

 equivalent harmonic motions, and the most important terms 

 retained, but there would be no necessary relation between the 

 number of these terms (sensible bright lines) and the number 

 of degrees of freedom of the molecule. H. \V. Watson. 



I S I UrK<: ihat th«; avr 

 tph-^f- are al any instant m 

 of cai'S in whitn two only <• 



NO. 13 14, vor. 51] 



The Horn Expedition to Central Australia. 



In your issue of September 27, 1S94, occurs a short notice of 

 the work of the Horn scientific expedition to Central Australia. 

 Reference is made therein to the discovery of a " new type of 

 marsupial " by Dr. Stirling. The animal in question was found 

 by Mr. South, a mounted trooper (or rather by his cat, who 

 brought it into the house), at Alice Springs. By him it was 

 presented on our arrival to Dr. Stirling, who had charge of 

 the anthropological work of the expedition, by whom it was 

 kindly handed on to me as officer in charge ol the zoological 

 department. The specimen was a male, and being desirous 

 of securing more, I stayed behind the party, and by aid of the 

 blacks procured two more specimens, both of them females. 

 The animal, which lives in holes amongst the rocks and stones, 

 is by no means common, as I h.id to ofler considerable quan- 

 tities of flour and tobacco to the blacks as a reward for its cap- 

 ture. .\fter a number of them h.ad been out hunting for several 

 days the total result was two more specimens, though as these 

 were females they formed a welcome addition to my zoological 

 collections. As the expression, "a new type of marsupial,'" 

 gives rise to too great expectations, it may be as well to state 

 that it is merely a new species of the genus Phascologale, dis- 

 tinguished, amongst other points, by itsremarl;ably fat tail and 

 by the nature of the striated pads on the soles of its feet. I 

 was able to make drawings of the animal alive, and on showin;; 

 these to the bkicks at Charlotte Waters, some 250 miles to the 

 south of Alice Springs, they were at once greeted with cries 

 of " Amperta, " the native name for the animal which they 

 look it to represent. Through the kindness of Mr. Byrne, the 

 head of Charlotte Waters Telegrapli Station, I have since 

 been provided with specimens of the ■'.\mperta," which on 

 examination turns out to be the rare form — only as yet, I be- 

 lieve, known from a single specimen — described by Kreflt 

 under the name of Chutoccrcus crislieanda, and subsequently 

 placed by Mr. Oldfield Thomas in the genus Phascologale. 



These two species, and a new one of the genus Saiinthopsis, 

 which we secured amongst the sand-hills near Lake .Xmadeus, are 

 the most important finds amongst the marsupials which, owing 

 to the country traversed by us bein^; in the main of a desert 

 description, were by no means plentiful. In this region animals 

 can only be secured in numbers after rain, an experience which, 

 during three and a half monlhs' wandering, did not fall to ourlot. 

 However, as one result of Mr. Horns generous action in 

 equipping the expedition, I hope to be able to give a fairly good 

 general account of the fauna of the central desert region of 

 Australia. Towards the close of the notice referred to, it is said 

 that there is some doubt as to the manner of publication of our 

 results. Mr. Horn's intention is, 1 believe, to issue a senaralt- 

 volume, the various parts of which will deal with the branches 

 of science represented by the members of the scientific staff as 

 follows : — 



Prof. R. Tate (chairman of the scientific staff), geology and 

 botany; Dr. IC. C. Stirling, anthropology; Mr. C. Winnecke 

 (leader of the expedition), surveying and meteorology ; Prof. 

 Baldwin Spencer, zoology ; Mr. J. A. Watt, geology and 

 mineralogy. 



The volume will contain an accurate map, compiled by Mr. 

 Winnecke, of the central district drained by the Finke river. 



