376 



NATURE 



\August 30, 1877 



at any required temperature. The resistances at different tem- 

 peratures under different durations of contact will, when tabulated, 

 represent a series of logarithms, the base of each system being 

 the ratio between the resistances for the same differences of 

 temperature, but corresponding to different periods of contact. 



l'"rom these facts, electrification phenomena are capable of 

 receiving a mathematical rendering, which must prove of great 

 use to telegraph engineers. 



If the temperature coefficient were constant for all and every 

 period of contact, we should obviously obtain a series of curves 

 with ordinates increasing in a constant ratio, which would mean 

 that the resistances did not diminish as we reach the higher 

 temperatures. Now as the temperature coefficients for increased 

 duration of contact diminish, the curves more nearly approach 

 each other as the temperatures are increased, which agrees with 

 the fact th.at electrification ratios are reduced less and less as the 

 resistance itself diminishes. A very curious result arifcs from 

 this treatment of the subject, which I have rot yet had sufficient 

 time to examine, but which may be nientionedjiere as probably it 

 may assist us to explain something of the nature of electrification. 

 To determine approximately the electrification ratio and conse- 

 quently the resistance at any required temperature and for any 

 period of contact, calculate first the temperature coefficient for 

 the required temperature, which is simply the expansion of the 

 ratio for 1° F. to that power corresponding to the degrees of 

 difference. Using this as the factor, extract the root of the ratio 

 for any two given periods of contact, this will give very nearly 

 the electrification ratio conesponding to the same period of 

 contact at the required temperature. 



It thus appears that electrification, which is an inseparable 

 property of all insulators follows some law of variation in which 

 the temperature coefficient of the insulator itself is a function. 



I hope to communicate to a future meeting the mathematical 

 development of the application of logarithmic functions to electri- 

 fication and thermal charges in insulating media. 



A'otcs on t/ie ]'oluiitcS oj Solutions, by J. A. Ewing and J. G. 

 MacGregor, D.Sc. — In a paper by the authors published in vol. 

 xxvii. of the Tramactions of the Royal Society of Edinburgh, 

 containing an account of experiments on the density and 

 electrical conductivity of certain saline solutions, notice is 

 directed to the fact that the density of very weak solutions of 

 sulphate of copper and sulphate of zinc is greater than it would 

 be on the hypothesis that the anhydrous part of the salt dissolves 

 without increase of volume in the whole of the water present, 

 including the water of crystaUisation. On the other hand the 

 density of comparatively strong solutions is less than this hypo- 

 thesis would make it. From this it follows that if a small 

 quantity of one of these salts in the anhydrous state were added 

 to water, it would cause contraction, while a larger quantity of 

 the salt would produce expansion. The amount of such contrac- 

 tion, however, as indicated by observations of density, was so 

 small, that the authors were unwilling to speak positively as to 

 its existence until they had applied a direct volumetric test. 

 They have now done so, with the result of confirming the deduc- 

 tion drawn from their earlier experiments. 



The apparatus consisted of a large bottle, 2744 c.cm. in 

 capacity, through the cork of which projected a vertical tube of 

 0'66 cm. in bore. The bottle, as well as a part of the tube, 

 was filled with distilled water, and the salt was introduced in 

 quantities of ten grammes at a time. The resulting change of 

 volume was shown by the rise or fall of liquid in the tube. In 

 order to eliminate the effect of variations of temperature, a 

 second precisely similar bottle and tube were prepared and filled 

 with water, and the two were placed together in a large tube 

 full of water. 



The second bottle acted as a thermometer, and the expansion 

 or contraction dr.e to the introduction of the salt into the first 

 bottle ^vas indicated by the difference between the changes of 

 level in the two tubes. After the introduction of each dose of 

 salt the bottle was rolled about for a lime, so as to secure 

 thorough diffusion and solution, and then an interval of at least 

 six hours elapsed before readings were taken, in order that llie 

 heat given out by the hydration of the salt might be dissipated. 



The following results have been obtained in the case of anhy- 

 drous sulphate of copper: — The maximum contraction occurs 

 when the proportion of anhydrous salt to water is about one to 

 fifty, and the amount of contraction is then 000043 of 'he original 

 volume of water. As more salt is added the solution begins to 

 expand, and with one part of salt to eighteen of water the 

 volume is equal to that of the water originally present. After 



this any further addition of salt produces expansion beyond the 

 original volume. The rate of expansion per unit quantity of 

 salt appears to increase continually, but at first it is negative. 



The above numbers are given subject to correction by more 

 elaborate experiments that are now going on. The authors 

 hope to extend the inquiry to other salts. They have already 

 examined the behaviour of anhydrous sulphate of soda, but vrith 

 that salt no contraction whatever has been observed ; the 

 solutions expand rapidly from the first. 



On Magnetic Induction as affectin;^ Ohsci'c'ations of the Inten- 

 sity of the Horizontal Component of the Earth's Magnetic Force, 

 by Charles Chambers, F.R.S., Superintendent of the Colaba 

 Observatory, Bombay. — The magnets used in observations of 

 deflection and vibration, which observations are necessarily made 

 in the field of the earth's magnetic force, are subject to the 

 inducing action of that force ; and it is the universal practice 

 of magnetic observatories, sanctioned by the most eminent 

 writers on terrestrial magnetism, to apply corrections on account 

 of induction both to the deflection and vibration observations. 

 The object of this communication is to advance theoretical 

 reasons, supported by experimental evidence, against the pro- 

 priety of the particular correction applied to the vibration 

 observation. This correction is based on the assumption that 

 the vibration magnet is susceptible of induction longitudinally 

 but not transversely or not so sensibly ; and the assumption 

 probably rests on what the writer regards as a false analogy 

 between a permanent magnet and an induced magnet. The 

 former, when removed from the inSuence of a strong mag- 

 netising action, remains a magnet by virtue of its own internal 

 forces, whilst the latter is a magnet by virtue of external forces 

 alone ; it does not therefore follow that because the power of a 

 permanent magnet, measured by its magnetic moment, cannot 

 be made by the same means nearly as great transversely as 

 longitudinally, therefore the same may be said of an induced 

 magnet. Indeed, in his treatment of the subject of the devia- 

 tions of the compass, Sir George Airy gives to each elemental 

 portion of a ship's iron as great a susceptibility to induction in 

 one direction as in another ; and in the more elaborate treatment 

 of the same subject, in which I'oisson's equations are taken as 

 expressing the fundamental conceptions of the theory, terms 

 representing transverse induction are still retained as of com- 

 parable magnitude in presence of others representing longitudinal 

 induction. 



Applying the Astronomer-Royal's theory to the paiticular case 

 of the vibration magnet, its induced magnetism becomes an 

 assemblage of elementary magnets, whose magnetic axes are all 

 parallel to the magnetic meridian, and which, since they sensibly 

 retain their parallelism to the meridian during the oscillation of 

 the magnet, give rise to no moment of restitution, hence, 

 according to this view, no correction would be required. 



According to Poisson's theory, the amount of the correction 

 is matter for experimental inquiry, and cannot be safely deter- 

 mined on () priori grounds. It may be objected, however, that 

 the swinging of a ship being a slow motion compared with the 

 oscillation of a magnet, the theory of the deviations of the 

 compass must be modified in its application to the case in 

 question ; and this is, no doubt, a correct view, for the theory 

 regards the inductive action as being, at every moment consi- 

 dered, sensibly carried to its limit of effectiveness ; whilst it is 

 not only conceivable, but doubtless the fact, that where, as with 

 the oscillating magnet, the motion is reversed every few seconds, 

 the transverse inductive action only partially approaches its 

 limit. On this account we should be prepared to expect then, 

 that even if the transverse induction were as great as the longi- 

 tudinal when time for full development of the induction was 

 allowed, it would be in defect in the case of the vibrating 

 magnet. 



In the years 1S73 and 1S74 — long before these views of the 

 subject of induction first occurred to the writer — he had had 

 made in Bombay a careful comparison o two Kew unifilar 

 magnetometers by means of practically contemporaneous ob- 

 servations. The result was to show a persistent difference in the 

 values of the horizontal force yielded by the two instruments, 

 far exceeding any probable errors of observation, and, after a 

 careful examination of each single observational quantity and of 

 each constant entering into the computations, the writer came to 

 the conclusion that no error of the magnitude of that in question 

 could have its source .anywhere but in connection with the 

 induction collections. The values obtained for the horizontal 

 force were, in British units of force— 



