144 



NA TURE 



[April i, 1909 



To study the conductivity of flame, it is convenient to 

 use a row of small Bunsen flames placed so that they 

 touch each other. I use a row of fifty flames burning 

 from quartz tubes i cm. apart. This gives a flame 50 cm. 

 long and about 10 cm. high. The quartz tubes Insulate 

 very well, so that a current can be passed along the flame 

 horizontally from one end to the other. 



When two parallel platinum electrodes immersed in the 

 flame are connected to a galvanometer and battery, it is 

 found that a measurable current is obtained. The relation 

 between the current (i) and the difference of potential (V) 

 between the electrodes is given by the equation 



V = Ai= + Bd.-, 



where A and B are constants, and d denotes the distance 

 between the electrodes. If i is small, say one or two 

 millimetres, the term Bdi is negligible (except when i is 

 very small), and we get \ = Ki^. In this case the current 

 is almost independent of the distance between the elec- 

 trodes. 



The reason for this peculiar relation between the current 

 and potential difference becomes apparent when the varia- 

 tion of the potential along the flame from one electrode to 

 the other is examined. An electrometer is connected to 

 two platinum wires, which are immersed in the flame, and 

 can be moved along horizontally between the electrodes. 

 Each wire takes up the potential of the flame at the point 

 where the wire is situated, so the deflection of the electro- 

 meter indicates the difference of potential between the two 

 points where the wires are put in. Suppose one wire is 

 allowed to touch the positive electrode and the other is 

 gradually moved along the flame from the positive to the 

 negative electrode. It is found that in the space between 

 the electrodes there is a small uniform potential gradient, 

 but near each electrode there is a comparatively sudden 

 drop in the potential. The drop near the negative elec- 

 trode is much larger than the drop near the positive 

 electrode. Thus nearly all the electromotive force of the 

 battery is used up close to the negative electrode. This 

 shows that nearly all the resistance offered by the flame 

 to the passage of the current is close to the negative 

 electrode. The positive ions in the flame move towards 

 the negative electrode and the negative ions towards the 

 positive electrode; in fact, the current is carried through 

 the flame by these two streams of ions. Hence, close to 

 the negative electrode, the current must be carried entirely 

 by positive ions moving towards it, and at the positive 

 electrode the current must be entirely carried by negative 

 ions. We find that the resistance near the negative elec- 

 trode is much greater than near the positive electrode, so 

 that we conclude that the negative ions carry the current 

 more easily than the positive ions. With a given electric 

 force, the negative ions move very much faster than the 

 positive ions. It has been shown experimentally that the 

 velocity of the negative ions is about 10,000 cm. per sec. 

 for one volt per cm., while that of the positive ions is 

 about 100 times smaller than this. 



In the flame away from the electrodes the electric force 

 is found to be proportional to the current, so that here the 

 flame obeys Ohm's law like a metallic conductor. Its 

 conductivity is about 10" times less than that of copper. 

 In the equation V = Ai^-l- Bdi', the term Bdi is the part of 

 the E.M.F. used up between the electrodes, so it is pro- 

 portional to the current and to the distance. Sir J. J. 

 Thomson has shown theoretically that the drop of potential 

 near the electrodes should be proportional to the square of 

 the current, as is found experimentally to be the case. 



The conductivity of a Bunsen flame may be compared 

 with the conductivity of liquids, such as water. In pure 

 water some of the molecules are dissociated into ions and 

 the water is a conductor, although only a poor one ; but 

 if a salt like sodium chloride is dissolved in the water, the 

 salt dissociates into ions almost completely, and the con- 

 ductivity is greatly increased. Suppose we hold a bead of 

 salt on a platinum wire in a flame, then the salt volatilises 

 and the flame is filled with its vapour, and, just as with 

 the water, the conductivity is enormously increased. 



With the long flame and an electrode at each end, we 



can try the effect on the current of putting salt in different 



parts of the flame between the electrodes. In this way it 



is easy to show that the current is practicallv unchanged, 



NO. 2057, ■VOL. 80] 



unless the salt vapour is put in close to the negative elec- 

 trode, but in that case it produces a very great increase in. 

 the current. This confirms the conclusion that nearly all 

 the resistance to the passage of the current is situated close 

 to the negative electrode. When the salt is put in any- 

 where it diminishes the resistance there to a small fraction 

 of its value, but it is only close to the negative electrode 

 that the diminution in the total resistance is appreciable. 

 If we measure the potential difference between two points 

 in the flame away from the electrodes, and then put salt 

 vapour in the flame between them, we find that the P.D. 

 drops to a small fraction of its value, although the current 

 is the same as before. This shows clearly that the salt 

 vapour greatly increases the conductivity wherever it is 

 put in. 



When some salt is put on the negative electrode, the 

 sudden drop in potential there almost disappears, and we 

 get a nearly uniform potential gradient from one electrode 

 to the other, so that now the resistance is nearly uniformly 

 distributed along the flame. If now salt vapour be put 

 in anywhere between the electrodes, the current is in- 

 creased. If, for example, we fill half the length of the 

 flame with salt vapour, we nearly double the current. 



When salt is put on one electrode, the flame can be 

 used as a rectifier for an alternating current, for when the 

 salted electrode is negative the resistance of the flame is 

 much smaller than when it is positive. 



Measurements have been made of the conductivities of 

 a number of alkali salt vapours in a current of air flowing 

 along a platinum tube heated in a gas furnace. An elec- 

 trode was fi.\ed along the axis of the tube, and the current 

 from it through the salt vapour to the surrounding tube 

 was measured with a galvanometer. It was found that at 

 temperatures above 1400° C, and with electromotive forces 

 of about 1000 volts, the current was proportional to the 

 amount of salt passing through the tube, and for different 

 salts in equal quantities inversely proportional to the 

 electrochemical equivalent of the salt. This shows that the 

 quantity of electricity per molecule of salt is the same for 

 all salts. It w'as also found that the quantity of electricity 

 carried per molecule was equal to that carried per molecule 

 when a solution of salt in water is electrolysed. It 

 appears, therefore, that the laws of electrolysis discovered 

 by Faraday for liquids apply also to salts in the state of 

 vapour. 



When a molecule of salt like sodium chloride dissociates 

 into two ions in water, the sodium atom forms the positive 

 ion and the chlorine atom the negative ion, and when a 

 current is passed through the solution the sodium is 

 attracted to the negative electrode and the chlorine goes 

 to the positive electrode. We might expect the same thing 

 to happen when a current is passed through the salt vapour 

 in a flame. If we put two wires in the flame, and put 

 some sodium salt on one, and then connect them to an 

 induction coil, and pass a discharge from the salted one to 

 the other, we find that the yellow sodium vapour appears 

 at it when it is the negative pole, but not when it is 

 positive. This shows that in the flame the positive ions 

 of the salt vapour contain the metal just as they do in 

 solutions. The negative ions, however, do not appear to 

 be the same in flames as in solutions. In flames the very 

 high velocity of the negative ions indicates that they are 

 the electrons the properties of which have been investigated 

 in vacuum tubes by Sir William Crookcs and Sir Joseph 

 Thomson. The positive ion, then, is an atom or molecule, 

 while the negative ion is an electron the mass of which is 

 several thousand times smaller. This is the explanation 

 of the fact that the negative ions move mo times more 

 quickly than the positive ions. 



HEAT-TRANSMISSION IN STEAM BOILERS.^ 

 A T the present tune the relations between the various 

 "^ factors that govern the flow of heat from a gaseous 

 fluid into a metal surface with which it is in contact remain 

 extremely obscure. 



The formulae in general use, which express in a con- 

 crete manner the views of engineers upon the subject, are 

 of a purely empirical character and without theoretical 

 1 Abstract of paper on *' Laws of Heat and Transmission deduced from 

 Experiment," by Prof. J. T. Nicolson, read before the lunior Institution of 

 Engineers. 



