November io, 1898] 



NA TURE 



processes. They may be allou-ed to diftuse freely into 

 one another, or may be separated by a porous par- 

 tition. In the latter case a considerable difference of 

 pressure may be produced between different parts of the 

 space containing them, and this difference of pressure 

 can be used to do work. The final condition in this case 

 is, of course, cooler than if the gases did no external work. 

 In the same way a solution diffusing into water may do 

 so without doing e.-cternal work, or it may do so by a 

 reversible process, through a semi-permeable diaphragm, 

 producing considerable differences of pressure, which may 

 be used to do work. The final condition in this latter 

 case would, of course, be cooler than in the former case 

 of inter-diffusion without doing external work. Now 

 whenever convection currents are produced, these are to 

 some extent reversible. We might put vanes into the 

 liquid to be moved by the currents and to do work out- 

 side the liquid, and by reversing this we would reverse 

 the convection currents. Hence any method of mixing 

 in which convection currents are produced, which do work 

 or produce heat outside the liquid, will necessarily produce 

 less heat in the liquid than a method of mixing in which 

 there are either no convection currents, or these produce 

 heat by viscous flow inside the liquid. 



That we can, at pleasure, either use the diffusion of 

 two gases into one another to do external work or not, 

 is really not different from the case of a single gas ex- 

 panding into a larger volume. We may do work by this 

 expansion and cool the gas, or we may allow the gas, as 

 in Joule's experiments, to expand into a larger volume 

 without doing external work, and in this case there is 

 only a very small change of temperature. 



In these cases it is a question of change of entropy in 

 the system, which can either be effected by an irre- 

 versible process m which no work is done, or by a variety 

 of other processes, more or less reversible, in which the 

 more reversible they are the more work can be done. In 

 the case of producing convection currents, or, in general, 

 of diftusion of a heavy fluid upwards into a lighter one, 

 the amount of heat produced would not be exactly the 

 same as if gravity were not acting : the centre of gravity 

 of the system is raised by diffusion. Now in Mr. 

 Griftlths's case, and in the case of diffusion currents 

 generally, this raising of the centre of gravity takes 

 place throughout part of the space considered by diffu- 

 sion, and the centre of gravity is continually falling down 

 again in the convection currents. Hence the work that 

 ■can be done by the convection currents is part of the 

 work that was done by diffusion against gravity. In the 

 case of diffusion without convection currents, we might 

 use the whole of this work done against gravity, by 

 which the centre of gravity of the system has been raised, 

 to do e.xternal work. If, for example, the containing 

 vessel were supported at its centre of gravity, in the un- 

 mixed condition, the centre of gravity would, after diffu- 

 sion, be above the point of support, and the vessel and 

 its contents might be arranged to turn round the support 

 doing work during the fall of the centre of gravity to 

 its original level. Another way of utilising the rise is to 

 I allow the fluid to flow into another broader vessel until 

 its centre of gravity has returned to the original level. 

 The thing to be specially observed is, that the amount by 

 ' which the centre of gravity is raised depends entirely 

 I upon the shape of the vessel. If it be tall, the centre of 

 graxity will be raised a great deal ; while if it be low, the 

 \ centre of gravity will' be only slightly raised. I5y 

 \ causing diffusion to take place in a tall thin vessel, 

 I the final temperature will be lower than in a 

 j broad low one, not on account of any superficial 

 I tensions, but on account of the work clone against 

 gjravity. In Mr. Griffiths's methods diffusion is con- 

 , tinually taking place along tall thin vessels, and con- 

 j vection currents lowering the centre of gravity again 

 by flow into broad ones. Gt:o. Fr.a.s. FitzGer.\ld. 



NO. 15 I 5, VOL. 59] 



THE EXPECTED METEORIC SHOWER. 



T^HE imminent return of the Leonids once more attracts 

 ■*- us to prepare for their observation and discuss their 

 phenomena. The circumstances this year will be much 

 more favourable, all round, than they were in 1897, but 

 our prospects of witnessing a really brilliant return appear 

 to be somewhat slender. No doubt, on the morning of 

 November 15, meteors will appear in sufficient abundance 

 to gratify moderate expectation, but the conditions 

 scarcely warrant the influence that we are to have a 

 grand display. We must wait until iSgg or 1900 to see 

 the shower at its best. In 1832 it is true Dawes saw 

 many astonishingly fine meteors ; and well he might, for 

 the parent comet of the Leonids was very near that 

 section of the orbit which the earth intersected in the 

 year named. In 1865 we passed through a region of the 

 stream some way in advance of the comet, for the latter 

 arrived at its descending node about two months after 

 the earth had crossed the point. There was nothing 

 deserving the title of a great meteoric shower on that 

 occasion. But there was certainly an unusual number 

 of fine shooting-stars, the majority of the objects 

 observed being as bright as, or brighter than, stars of the 

 first magnitude. At Greenwich it was estimated that 

 more than 1000 meteors must have been visible on the 

 morning of November 13. Mr. Knott, observing at 

 Cuckfield in Sussex, estimated the number as more than 

 one per minute for two observers. According to some other 

 accounts the richness of the display far exceeded this, 

 for a captain of a British ship, near the West Indies, 

 wrote to say that the heavens were in a blaze with 

 shooting-stars from 8 p.m. on November 12 to 5 next 

 morning. But accounts of the latter description are 

 often exaggerated, and it is always unsafe to draw any 

 definite conclusions from them. 



At the approaching return the earth crosses the 

 meteoric orbit still further in front of the comet than it 

 did in 1S65. In fact the comet will have five or six 

 months' journey to run at its highest rate of speed before 

 it reaches its descending node. This is not allowing for 

 any perturbations which the comet has experienced 

 since 1866, and there is no doubt that some serious 

 disturbances have been introduced, particularly, by 

 Saturn and Jupiter. 



It seems that in July 1S95, the comet approached to 

 within 45 millions of miles of Saturn, and though the 

 former has not passed so near as this to Jupiter, both 

 planets have exercised a very appreciable influence both 

 on the comet and its associated meteoric stream. Dr. 

 Berberich gives these conclusions in an important paper 

 published in Ast. Nach., 3526, and states as a result of 

 his investigation that the meteor shower will appear 21 

 hours late in 1898 and 26 hours behind time in iSgg. 

 The comet of Tempel (1866 I.) is not, according to Dr. 

 Berberich, likely to be observed at the ensuing return to 

 perihelion, as it will present itself under unfavourable 

 conditions. Dr. Berberich's results are interesting as 

 showing the necessity for expecting the meteors on the 

 mornings of the 15th and l6th, rather than on earlier 

 dates. His conclusions seem strengthened by the fact 

 that last year a pretty strong shower of Leonids was 

 witnessed just before sunrise on the morning of the 15th, 

 whereas very few were seen on the previous morning. 



Under all the circumstances a very rich shower can 

 hardly be expected. Our historical records do not 

 warrant the assumption that the section of the orbit in 

 the van of the comet is thickly strewn with meteoric 

 particles. In the comet's wake, for an enormous distance, 

 the material appears to be densely distributed. This was 

 sufficiently attested by the succession of three brilliant 

 displays of 1866, 1867 and 1868. 



Meteoric and cometary phenomena are, however, 

 somewhat unstable in character, and certainly variable 



