Oct. 7, 1880] 



NATURE 



535 



picking up another at the next station, each carriage thus slowly 

 shifting round the line. 



But any such plan would entail a fresh build of carriages ; and 

 for discontinuous carriages a plan nearly as good would be to run 

 a railway omnibus on the rails, with a small 6 or 8 h.p. engine all 

 in one. This would be stopped anywhere between stations, at 

 crossings, farmhouses, and hamlets along the line, and would 

 serve the peasantry for going shopping, beside taking up baskets 

 of garden produce. Passengers going a long journey would 

 change at a main station and join the ordinaiy train, which 

 would only stop about every hour at the ends of forty or fifty 

 miles' stages. Country lines only running a train every two hours 

 or so would be easily worked thus, the 'bus being shunted by 

 telegraph if necessary, and the line signalled clear as usual. With 

 double lines the 'bus would run on the goods line. 



Bromley, Kent W. M. F. Fetrie 



A NEW KIND OF ELECTRIC REPULSION"- 



DR. GOLDSTEIN has devoted a good part of the last 

 ten years to an investigation of the discharge of 

 electricity through gases, and amongst the many pheno- 

 mena which he has brought to light, the one described 

 in a memoir published in a separate form is not the least 

 interesting and important. The facts may be stated in a 

 few words : A negative electrode exerts a strong repulsion 

 on the rays of the glow proceeding from itself or from 

 another negative electrode. Before describing the e.xperi- 

 ments proving this statement, and the laws by which this 

 phenomenon is regulated, we shall follow Dr. Goldstein 

 in reminding the reader of a few facts connected with a 

 discharge of electricity through gases which he will have 

 to bear in mind. 



It is well known that the negative electrode in a gas, 

 for which Faraday's name of cathode may be conveniently 

 used, is surrounded with a glow which expands as the 

 pressure of the gas is reduced. We are able to distinguish 

 four layers in this gas, though three of them only are 

 easily recognised. As a first approximation we inay 

 assume the outline of these layers to be parallel to the 

 outline of the electrode, though, as we shall have to men- 

 tion, Dr. Goldstein has shown that this is not strictly correct. 



The layer of the negative glow adjacent to the cathode 

 is luminous, and shines in air with a yellowish-red tint. 

 This first layer is surrounded by a second layer, which is 

 very little luminous. This is the dark space mentioned by 

 Mr. Crookes ; but, as Dr. Goldstein shows, it is not 

 entirely dark, but has in air a bluish tint. We next come 

 to the third and fourth layers, which may very well be 

 taken as one, and which are more generally designated by 

 means of the teiTn, negative glow. They form the outer 

 boundary of the luminosity surrounding the cathode. If 

 the pressure of the gas is sufficiently reduced to enable 

 the glow to touch the glass, it becomes phosphorescent, 

 and only the layer of the gas immediately touching the 

 glass causes the phosphorescence. The phosphorescence 

 gets stronger as exhaustion proceeds ; at the same time 

 the luminosity of the glow gets weaker. The appearance 

 and extension of the glow does not depend on the position 

 of the anode, while the luminous positive discharge varies 

 very much with the relative position of the electrodes, and 

 can be made to disappear altogether by bringing the 

 electrodes sufficiently near. 



Already Pliicker, and especially Hittorf, have come to 

 the conclusion that the negative glow is propagated in 

 rectilinear rays from the cathode, and it can further be 

 shown that the direction of propagation is generally in a 

 direction nearly normal to the surface of the cathode. Dr. 

 Goldstein draws a distinction between such elements of 

 the cathode which lie near the edge, if the surface of the 

 cathode has edges, and elements which are removed from 

 the edge. While those elements not near an edge only 

 send out rays within a cone of narrow aperture in a 



•_" A New Kind of Electric Repulsion," by Dr. E. Goldstein. (Berlin : 

 Julius Springer, 1880.) 



normal direction, the edges send out rays in all direc- 

 tions. This difference in the behaviour of different 

 elements of the same surface is, it appears to us, well 

 explained by Dr. Goldstein's discovery of a repiilsion 

 between the electrode and a ray proceeding from the 

 cathode. A little consideration will show that this repul- 

 sion will, whenever cylindrical or plane electrodes are 

 used, be in a nearly normal direction for any part of the 

 surface which is sufficiently removed from the edge, while 

 near the edge the resultant repulsion will be away from 

 the surface and from the greater angle with the normal 

 the nearer the ray is to the edge. This would prove of 

 course that the repulsion is not an electrostatic one, for 

 in that case it would always be at right angles to the 

 surface. If the exhaustion is such that the glass becomes 

 phosphorescent, the phosphorescence, being produced by 

 the rays proceeding from the cathode, it is clear, wiU form 

 a luminous ribbon surrounding the electrode, which is a 

 little larger than the electrode. 



If now a solid body is introduced between the cathode 

 and the glass inclosure, a shadow of this body will appear 

 in the phosphorescent light on the glass ; the formation 

 of the shadow is a direct consequence of the rectilinear 

 propagation of the rays. 



We now proceed to describe Dr. Goldstein's experi- 

 ment in its simplest form. 



In a cylindrical vessel two parallel electrodes of equal 

 length are introduced at one end, while the other end 

 contains a third electrode which shall always form the 

 anode. Let the pressure be such that phosphorescence 

 appears, and let only one of the two parallel electrodes 

 be connected with the negative pole of the coil, while the 

 other is insulated. A shadow of this insulated wire 

 is seen in the phosphorescent light on the glass. Now 

 let the insulated wire be brought into metallic contact 

 with the other electrode, and the whole appearance will 

 change. In the phosphorescent light of the glass we 

 shall see two dark surfaces of equal size and shape, and 

 with distinctly marked edges. The two dark surfaces 

 are situated in such a way that a plane which passes 

 through the electrodes cuts them into two equal halves. 

 They are partly bounded by straight lines, partly by two 

 semicircular arcs. 



The parts formed by straight lines are parallel to the 

 electrodes, and of equal length ; these straight lines are 

 joined at the lower end, that is, at the free end of the 

 electrodes, by means of a half circle, which is partially 

 repeated at the upper end ; but where the electrodes are 

 sealed into the glass the curve is interrupted. The dark 

 surfaces are bordered by a bright line of light. It will 

 facilitate the understanding of the position and shape of 

 these dark surfaces if we mention already here that they 

 are such as would be produced if the rays emanating 

 from each electrode, and propagated in a normal direc- 

 tion from it, suffer a repulsion and consequent deflection 

 in the neighbourhood of the other electrode, so that the 

 dark space is formed by the absence of the phosphores- 

 cent light which would be produced by the rays coming 

 from the farther cathode. 



We cannot here give the further description of shape 

 and the measurement of the size of these dark surfaces, 

 but at once describe their properties. In the first place 

 the size and shape are altogether independent of the 

 position, form, and size of the positive electrode. The 

 relative position of the two cathodes, on the other hand, 

 materially affects their behaviour ; and Dr. Goldstein 

 gives their shape, for instance, if, instead of being 

 parallel, they are at right angles to each other, either in 

 the same plane or one in front of the other. We have 

 already stated that in the case of parallel electrodes the 

 parts of the outline forming straight lines are of equal 

 length with the electrodes, and hence the length of these 

 dark surfaces increases with the length of the electrodes, 

 but the breadth and half-circle joining the straight lines 



