570 



NATURE 



[Oct. 13, 1 88 1 



in our knowledge of the nature and modus operandi of the dis- 

 charge, we ou^Iit to be in a position to modify the discharge, so 

 as to compare it under different circumstances. The methods 

 hitherto usually employed for this object have been an alteration 

 in the gas used, an alteration of the pressure, and a diversity of 

 figure in the tube or in its terminals. Of the general character 

 of the changes due to such alterations we have already seen 

 something. But, besides these alterations, which are of a struc- 

 tural or instrumental character, it is also desirable to operate on 

 adischirge actually z« fi-attsilii. One of the methods, in fact 

 the only one, employed until lately is that of the magnet, which 

 was used as long ago as the time of Grove, PUicker, and other 

 early experimenters. It is well known that a magnet will dis- 

 place a movable conductor when carrying a current, according 

 to laws established by Ampere. The same is true, in general 

 terms, with respect to a discharge traversing an exhausted tube. 



Thu-, in the tube now before you, you will see that when one 

 pole of a magnet is presented to the tube, the discharge is 

 thrown to one side of the tube ; and when the other pole is so 

 presented the discharge is thrown to the other side. These two 

 main features, however, very inadequately describe the actions of 

 a magnet, which in fact operates separately not only upon each 

 stria as a unit, but even upon the various parts of a stria in such 

 a way as to deform it a. well as to displace it. But they involve 

 the main characteristics of the magnetic action, and must suffice 

 on the present occasion to show that in the magnet we have a 

 powerful instrument for examining the properties and functions 

 of the discharge, even (if the term may be penuitted) in the 

 living specimen. 



The other principal mode of operating on the discharge con- 

 sists in reducing it to what has been called the "sensitive state"; 

 i.e. to a state in which the position of the luminosity is affected 

 by the approach of a conductor to the tube. For the details of 

 an experimental investigation into the phenomena of this .state, 

 and a discussion of the conclusions that may be draw n therefrom, 

 the reader is referred to the Philosophical Transaclions of 1879 

 and 1880. But the following remarks may serve to convey some 

 notion of the method and its issues. Sensitiveness is produced 

 by breaking the circuit with a short intei-val of air, or, as it is 

 usually described, by interposing an "air-spark " in one branch 

 of the circuit, viz. either that leading from the positive, or in 

 that leading from the negative, terminal of the machine to the 

 tube, or by otherwise rendering the discharge intermittent. The 

 effect of this is to discharge the electricity discontinuously, so 

 that from time to time there passes into the tube a comparatively 

 large quantity of electricity at a higher tension than wriuld other- 

 wise be the case. By this means the gas in the inter! ir of the 

 tube, or perhaps the ulterior surface of the tube itself, becomes 

 mom-ntarily charged with electricity, thus creating an electric 

 ten-ion, which may be discharged or "relieved " by a displace- 

 ment in the electricity on a conductor brought near, or in contact 

 with, the tube. This causes or permits a discharge from the 

 interior of the tube itself of the electricity of an opposite kind to 

 th.it with which the tube itself is charged. If the air spark is 

 on the positive side the charge on the tube is positive, and the 

 relief negative, and vice versa. From this it follows, as might 

 have been expected, that the effect of the relief on the visible 

 discharge is different in the two case . In each case the part of 

 the inner surface of the tube nearest to the conductor acts as a 

 quasi terminal. As a general rule, with a positive air-spack, the 

 relief, being negative, tends to produce a dark space, and thereby 

 gives the appearance of a repulsion of the luminous column. 

 With a negative air-spark the relief, being positive, tends to pro- 

 duce a stria ; it thereby cau-es luminosity, and gives an appearance 

 of attraction of the luminous column. 



The appearance of the luminous column when produced under 

 the action of an air-spark is usually amorphous or unstratified, 

 although this is not always the case. In the case of a positive 

 air-spark the column is more or less constricted and confined to 

 the central part of the tube; in the case of a negative air-spark 

 it is more diffused, and usually fills the whole diameter of the 

 tube. This is doubtless due to a gradual discharge from the 

 sides of the tube, and is in accordance with what has been said 

 above. 



Under suitable circumstances this relief discharge may be 

 made to bring out artificial striae from an amorphous discharge, 

 the position of the striae depending upon the character of the air- 

 spark u-ed. The positions occupi-rd by striae in the one case 

 will be occupied by dark space- in the other, and lice versa. 



The facts here adduced, together with many others based upon 



a long series of experiments, all tend to the co iclusion that, 

 whatever the number or form of the striae in a stratified column, 

 each stria is to be regarded as a physical unit, and that in each 

 unit we have represented all the elements of a complete dis- 

 charge. The form of each stria is in every cise determined by 

 that of its immediate predece-sir, reckoned from the negative 

 end. This may be verified, among other ways, by observing 

 the form of the successive stri^ when distorted by the influence 

 of a magnetic field. The same research has further established 

 the fact that the negative glow and the haze behind it, which 

 terminates in what is usually kn jwn as the neJative dark space, 

 is a stria turnel as it were inside out by the influence, the shape, 

 and the character of the negative terminal. From the mode in 

 which this stria is connected with the negative terminal it has 

 been called the "anchored stria." 



The relief effects, may, however, be produced equally well by 

 connecting a point on the surface of the tube with the opposite, 

 or non-air-spark terminal, instead of with earth. By this means 

 we supply a charge of electricity of the opposite name to that 

 w ith which the tube has been charged, and obtain a result of a 

 similar character to that of ordinary relief. 



Alongside of the relief effects above mentioned there is also 

 a sy-tem of what we have termed "special effect^,'' which latter 

 are converse to the former, each to each. These are produced 

 by connecting a point on the outride of the tube with the air- 

 spark terminal itself. The special effect with a positive air- 

 spark is equivalent to a relief effect with a negative air-spark, 

 and vice versa. 



Lastly, all these effects may be produced by mean-, of im- 

 pulsive discharges to the outside of the tube from an independent 

 source of electricity, such as a second Holtz machine. And, 

 mutatis mutandis, the corresponding effects may be produced by 

 this method even on a non-sensitive discharge. This completes 

 the entire cycle of phenomena due to impulsive action ah extra. 



The character of these effects being known once for all, this 

 impulsive action may be used as a test of the nature of a dis- 

 charge {i.e. whether positive or negative) passing through a given 

 tube. For example, we may experimentally verify in the case 

 of a coil di-charge what might have been anticipated on the 

 principles now established. Such a discharge is in fact equally 

 intennittent from both ends. There is no reason why either 

 terminal should be regarded as the air-spark terminal rather than 

 the other. Hence we might expect that the discharge would be 

 positive through about one-half of the tube, and negative through 

 the remainder, with a neutral zone between them. And such 

 proves to be the case. But more than this, if we attach a small 

 conden-er to either terminal of the tube, so as to tone down the 

 impulsiveness of the discharge at that end, we can thereby alter 

 the proportions of the positive and the negative parts of the 

 di-charge and shift the position of the neutral zone at will. 



The distinctive character which it is thus po silile to convey to 

 the whole, or to the two parts of one and the same discharge, 

 naturally leads us to examine whether it be not possible entirely 

 to separate one from the other, and to produce what may be 

 called a uni]3olar discharge. And this in fact may be done ; by 

 connecting the one terminal of the tube through an air-spark to 

 one br.mch of the circuit, and by leaving the other disconnected, 

 we may pr duce a discharge which, having plunged blindly into 

 the tube, and finding no res;ionse from the other end, returns 

 upon i'self, and finds exit by the way by which it came. The uni- 

 polar discharge is essentially intermittent, and therefore sensitive ; 

 the positive is conical in form and tapering towards its end ; the 

 negative is broad, and, so far as it extends, it fills the entire 

 width of the tube. Lastly, two unipolar discharges of the same 

 name can be produced in the same tube ; they repel one another, 

 and each returns like a single one. 



From these experiments we conclude that the independence of 

 the discharge from each terminal is so complete that we can at 

 will cause di-charges from the two terminals to be equal in 

 intensity but opposite in sign (as in the case of the coil), 

 or of any degree of inequality (as in the case of the coil with a 

 small condenser). Or we can cause the discharge to be from 

 one terminal only, the other terminal acting merely receptively 

 (as in the case of the air-spark discharge) ; or we can cause the 

 di chirge to pass from one terminal only and return to it, the 

 other terminal not taking any part in the discharge ; or, finally, 

 we can make the two terminals pour forth independent dis- 

 charges of the same name, each of which passes back through the 

 terminal whence it came. 



One of the most important consequences which follows from 



