August 31, 191 1] 



NATURE 



281 



rocks, does nothing " to help the prospector, is made 

 in forgetfulness ol the exhaustive and well-illustrated 

 "Catalogue of the Glossopteris Flora" which was 

 published by the Trustees of the British Museum six 

 years ago, partly with the view of helping the develop- 

 ment of the coal-fields in India and the southern 

 hemisphere, where the handbook is extensively used. 



S. H. BURBURY, F.R.S. 

 DY the death of Mr. Samuel Hawksley Burbury, 

 -D on August lb, at eighty years of age, we have 

 lost an ardent worker in the domain of mathematical 

 physics who did much to elucidate the mysteries of 

 several problems in molecular dynamics. Mr. Bur- 

 bury was the son of Mr. Samuel Burbury, of Leam- 

 ington, and was born at Kenilworth in May, 1831. 

 He was educated at Shrewsbury and St. John's 

 College, Cambridge ; he was Craven University 

 scholar, Chancellor's medallist, Browne medallist, 

 twice Porson prizeman, and fifteenth Wrangler and 

 second in classical tripos, 1854. He was called to the 

 Bar at Lincoln's Inn in 185b, but his new profession 

 did not prevent him from continuing his mathematical 

 studies, and he thus became one of the few workers 

 in this country who have produced original mathe- 

 matical investigations while engaged in duties other 

 than that of a mathematical teacher. 



Much of Mr. Burbury's work was done in collabora- 

 tion with the late Rev*. H. W. Watson, F.R.S., with 

 whom he shared the joint authorship of treatises on 

 "The Application of Generalised Coordinates to the 

 Dynamics of a Material System" (1879) and "The 

 Mathematical Theory of Electricity" (1883-5), ' n me 

 latter of which the authors endeavoured to place 

 electrostatics and electromagnetism on a more formal 

 basis than had been done by Clerk Maxwell in his 

 original treatise. It is perhaps a pity that this book 

 appeared at a time when experimental developments 

 were beginning to break down many of our precon- 

 ceived electrical theories, and, so far as we can gather, 

 Watson and Burbury's treatise is not studied so much 

 as it deserves. It affords a fairly satisfactory repre- 

 sentation of electrical phenomena as known at the 

 time, but, of course, every mathematical or dynamical 

 theory of physical phenomena can only be regarded as 

 a scheme for coordinating results of experiments in 

 the simplest form; and with the discovery of new facts 

 every such scheme is liable to be found inadequate 

 when the necessity arises of superseding it by a more 

 comprehensive scheme. It is thus probable that at the 

 present time there is no scheme which represents our 

 existing knowledge of electrical phenomena much 

 better than Watson and Burbury's represented our 

 knowledge at the time it was written. 



But, like his friend Watson, Burbury seems to have 

 chosen as his favourite study that branch of molecular 

 dynamics in which Boltzmann occupied a central posi- 

 tion. Burbury and Boltzmann were certainly in con- 

 stant correspondence with each other, and many of 

 Boltzmann's papers were evidently the result of Bur- 

 bury's criticisms. Perhaps Burbury's training as a 

 barrister gave him special qualifications for playing 

 th= rule of critic; at any rate, if there was a weak 

 point in any argument Burbury would certainly find 

 it out in a very short time. A great amount of time 

 was spent in examining Boltzmann's "Minimum 

 Theorem," according to which an assemblage of 

 molecules representing on the kinetic theory a per- 

 fect gas, tends to assume the distribution commonly 

 known as "Maxwell's Law." In the proof of this 

 theorem the question of reversibility plays an im- 

 portant part, and it cannot be said that the introduc- 

 tion of probability considerations altogether overcame 

 \"0. 2183, VOL. 87] 



the ditnculty 01 accounting tor an irreversible pheno- 

 menon by means of a system the elements of which 

 were subject to the equations of motion of reversible 

 dynamics. 



In his "Kinetic Theory of Gases" (1899), Burbury 

 advanced a novel theory as to the distribution of 

 velocities in a medium the molecules of which were 

 too close together to satisfy the fundamental hypo- 

 theses involved in the proof of Maxwell's law. 

 According to Burbury, the velocities of neighbouring 

 molecules would become correlated, the probability 

 factor involving, besides the usual exponential of the 

 energy, the exponential of the vector product of the 

 velocities of pairs of molecules. Burbury further 

 offered, on this hypothesis, a tentative explanation of 

 liquefaction. According to Burbury, Maxwell's law 

 would thus become inapplicable to a dense gas. On 

 the other hand, if applied to rare gases, it leads to 

 the conclusion that helium and hydrogen cannot 

 escape from the earth's atmosphere, a conclusion 

 which the late Dr. Stoney stated was not in agree- 

 ment with his observations. Thus the kinetic theory 

 of gases affords another instance of a scheme which 

 covered our knowledge of physical phenomena at one 

 time but no longer does so. To overcome this diffi- 

 culty we are now resuscitating kinetic theories, but 

 employing them on a smaller scale than before — to 

 electrons instead of molecules. 



Another question which occupied Burbury, especially 

 during recent years, was the loss of availability which 

 occurs when two gases mix by diffusion of constant 

 temperature. As Burbury argued, if this happens 

 when different kinds of molecules mix by diffusion, 

 the same thing should be true when molecules of the 

 same kind mix by diffusion. Burbury's views on this 

 subject were stated in Science Progress a few years 

 ago. This problem again gave Burbury scope for 

 his critical mind. It cannot be said, however, that 

 he, or, indeed, anyone else, has succeeded in giving a 

 reason why the total entropy of a litre of one gas and 

 a litre of a second gas is equal to the entropy of the 

 mixture when its volume is one litre, while if two litres 

 of the same gas at the same pressure are allowed to 

 mix, the sum of their entropies is equal to the entropy 

 of the mixture when its volume is two litres. What 

 Burbury really showed by his arguments was that the 

 truth or otherwise of these statements can only be 

 tested by experimental evidence. 



At the present time the study of mathematical 

 physics is rather out of fashion in this country, and 

 it is not unusual to deprecate this study on the ground 

 that it frequently fails to account for the results of 

 observation. Is not this failure one of its most valu- 

 able features? Whenever a new physical phenomena 

 is discovered, plenty of people are ready enough to 

 invoke molecules, the aether or electrons, to account, 

 to.- it, and to talk about the motions of these which 

 give rise to the observed phenomenon. The mathe- 

 matical physicist comes along and says, "Very well; 

 then let us write down the equations of motion and 

 see if the reasoning works out correctly." He obtains 

 a result which does not account for experimental 

 conclusions. Which is wrong? Not the mathemati- 

 cian who has merely attempted to place the reasoning 

 on an exact basis, but the unmathematical physicist 

 who has endowed his ether molecules, his ether, or .his 

 electrons with properties that are incompatible with 

 each other. In Mr. Burbury we have had a worke^ 

 who never flinched from the task of following a line 

 of argument up to its logical conclusions, however 

 much these might run counter to orthodox views. He 

 would drive his opponent from one stronghold to an- 

 other, the controversy' being conducted in the most 

 friendlv way the whole time. The contest would fre. 



