January 5, 1899] 



NA TURE 



223 



j.iarallel nioliun ; whereas, it is entangled ami bruken up in the 

 porti'tn where the sinuous motion of the water occurs.' 



" It will be observed that this explanation is tea certain extent 

 the same as that subsequently given by Prof. Osborne Reynolds 

 in the letter to NATURE, from which I have quoted, and to 

 which I refer under heading (2). 



" The experiments of Prof. Reynolds, which he cites himself, 

 are entirely different from those of which I have given an account, 

 and ^eem to me to have but very little bearing upon the be- 

 haviciur of water in the conditions under which my experiments 

 were conducted. 



" With regard to the differences of state in the water in the 

 light band, I will not trespass on your time with a repetition of 

 the arsiuments which I have already published, and which have 

 not hitherto been answered, but I would point out that the 

 experiments you have seen to-night form a most striking method 

 of putting the theory to the test by actually placing liquid under 

 the condition of the thin border, and in obtaining when this is 

 <lone. results which are absolutely different from those which 

 were given by the thick film or sheet of water. This is one of 

 the strongest possible arguments in favour of the views which I 

 ha\e advanced, inasmuch as the results of the experiment seem 

 to have surprised some of the greatest authorities on the subject. 



" Coming to (3), I would first remark, though it has never 

 been distinctly claimed, that the water charged with air actually 

 represented real stream-line motion, in my first paper it was 

 stated that ' there was no difference whether the smallest quantity 

 of air was present, or so large a quantity as to render the slide 

 opaque, as the general behaviour of the flow of water was 

 practically unaffected by the presence of the air.' 



'■ Though this remark only applied to the general nature of 

 the results obtained, the experiments brought out very strongly, 

 I believe for the first time, various points which had long been 

 known to the naval architect, but which had not been actually 

 made visible to the eye. Prof. Osborne Reynolds, however, 

 states that ' air bubbles are the most misleading bodies that can 

 possibly be chosen to indicate the lines of motion in a fluid in 

 sinuous motion.' This remark led me to consider the very 

 approj'riate experiment of trying the effect of first taking the 

 air-charged water upwards and then taking it downwards under 

 the same conditions. Inasmuch as the air in one case was try- 

 ing to rise upwards through the water, which is moving in the 

 same direction, whereas in the other case it is trying to move 

 in the opposite direction, it is perfectly certain that if there w.as 

 much difference in the flow owing to the presence of air, it 

 would become marked under those circumstances. 



"The two photographs. Figs, i and 2, represent the re- 

 sults of this experiment, and I venture to think that no 

 one would be able to tell from the photographs themselves 

 where the air-charged water was flowing upwards or where it 

 was flowing downwards. This, I think, shows in a very 

 striking way the comparatively small effect which the presence 

 of air under siiitabU conditions, has upon the flow of the water, 

 and it is quite contrary to that which Prof. Osborne Reynolds 

 seems to anticipate." 



NO. 1523, VOL. 59] 



The truth of the whole matter appears to be this, that as far 

 as Prof. Osborne Reynolds has dealt with the behaviour of air 

 in water, it has been under conditions represented by plates 

 suddenly immersed or moving through still water, in which case 

 air has been in a state of bubbles of large size, whereas in my 

 own experiments, the air has been broken up into very minute 

 bubbles. The behaviour in the two cases has been as different as 

 that of a mass of water, say from a bucket, thrown through the air, 

 and the finely-divided particles of moisture in a fog or mist. The 

 effect of viscosity makes all the difference in the two cases. As, 

 however, I have never had an opportunity of seeing Prof 

 Osborne Reynolds's experiments, and as he has never witnessed 

 mine, perhaps his disagreement with me is, after all, only 

 another illustration of the old fable of the chameleon. 



H. S. Hele-Shaw. 



University College, Liverpool, December 21, 189S. 



Etherion. a New Gas.' 



So.ME months ago the discovery of a new gas, by Mr. C. 

 Brush, was announced by nearly all the scientific periodicals of 

 the world, which was said to be endowed with quite extraordinary 

 properties ; as, for instance, density one ten- thousandth of 

 hydrogen, molecular velocity and heat conductivity hundred 

 times that of hydrogen ! 



It seemed strange to me that such tremendous assumptions 

 should be based on no more convincing arguments than the 

 experiments (Mr. Brush's) reported therein, on the relative in- 

 crease of condition of heat in rarefied air, when glass powder, 

 contained in the same vessel, was being heated, and I was 

 waiting eagerly for the publication of the original account (Mr. 

 Brush's), since it struck me that all observed phenomena could 

 be explained by the well-known properties of water vapour. 



I was glad to learn afterwards, from a paper in the Chemical 

 A'ews (November 4, p. 221), that this is also the opinion of Sir 

 William Crookes, undoubtedly the greatest authority in this 

 kind of research. 



Now the original paper (Mr. Brush's) appeared in Science 

 for October 14. It has been already the subject of a severe 

 criticism by "A Physicist" in the Chemical Netvs (December 2, 

 p. 277), as it does not contain indeed any further argument for 

 Mr. Brush's hypothesis. I do not think it superfluous, however, 

 to warn in Nature, too, against an excessive credulity in this 

 matter, and to point at some facts not yet emphasised suf- 

 ciently by other sceptics. 



First I must mention, for the sake of those readers who are 

 not sufficiently acquainted with these things, that the conduction 

 of heat by gases, when not disturbed by convection currents, is 

 independent of the pressure, until this comes down to several 

 millimetres of mercury ; then it begins to decrease, at first very 

 slo wly, then faster, until it becomes nearly proportional to the 

 pressure, at the highest rarefactions ; and the differences of 

 conduction in various gases, very marked at higher pressures, 

 are much less at pressures of several millionths of an 

 atmosphere. 



These facts, which are in strictest accordance with the 

 kinetic theory of gases, as I have shown in the Phil. Mag. 

 (August 1S9S), have been investigated besides by other ob- 

 servers very carefully, also by Mr. Brush himself, and form the 

 object of a very interesting paper of his in the PItil. Mag. 

 ( lanuary). 



Mr. Brush found that the conduction of heat is increased 

 very much at high exhaustions in respect to other gases at 

 coi responding pressures, when glass powder is heated (but re- 

 maining always much smaller than at normal pressures), and he 

 infers from this that a gas of enormous conductivity is given 

 off by heated glass powder. 



Now to me the main point seems to be — how did Mr. Brush 

 measure these low pressures? By an improved form of the 

 Macleod gauge, which seems to be very suitable for dry gas, 

 but of course, like every Macleod gauge, is quite unfit for ga,ses 

 where moisture or other condensable vapours are present, since 

 then it indicates only the partial pres.sure of the not condensable 

 gas. He did not use any drying agents, since they absorbed the 

 gas in question, therefore the indications of the gauge are of no 

 value whatever as to the total gas pressure. 



When Mr. Brush measured a gas pressure of 0*38 millionth, 

 and found the conduction to be about forty times that of dry 

 air at this pressure, he may have had in reality a pressure of 

 038 millionth of air, and besides a pressure of 20 millionth of 



