June 30, 1898] 



NATURE 



215 



ments were made on nitrogen prepared from ammonium nitrite, 

 whence it appears that the presence of argon does not affect the 

 kathode fall. 



The value found for nitric oxide is of very doubtful validity. 

 The appearance of the discharge showed that the gas is rapidly 

 decomposed, and the fact that the kathode fall is nearly the 

 same in nitric oxide, air and nitrogen with a trace of oxygen, 

 points to the oxygen being the sole carrier of the electricity in 

 each case. 



If we leave nitric oxide out of account, it appears that the 

 kathode fall is an additive quantity, and hence a property of the 

 atoms rather than of the molecules. 



Assuming that the conduction in gases is electrolytic, the 

 analogy of the electrolysis of liquids suggests the possibility that 

 the kathode fall is a measure of the heat of dissociation of the 

 pas. If this be the case, the experiments might be taken to sup- 

 ! port J. J. Thomson's view that the carriers of the current are 

 provided by the disintegration of the atoms into much smaller 

 pa rticles. 



June 9. — "Experiments on Aneroid Barometers at Kew Ob- 

 servatory and their Discussion." By C. Chree, Sc.D., LL.D., 

 F.R.S. 



The paper deals with two species of data. The first consists 

 of particulars derived from the records of Kew Observatory as 

 to the errors in aneroid barometers subjected to the ordinary 

 Kew test, which consists in lowering the pressure to which the 

 aneroid is exposed inch by inch to the lowest point required, 

 and raiding the pressure in a corresponding way to its original 

 value. Readings are taken at each inch of pressure during both 

 the fall and the recovery, and a table of corrections is obtained 

 by reference to a mercury gauge. 



The second group of data are the results of special experi- 

 ments made at Kew Observatory during the last three years. 



The aneroid is an instrument exhibiting elastic after-effect. 

 When pressure is lowered and then maintained constant, the 

 reading continues to fall, and when pressure is restored to its 

 (liginal value, the aneroid reads at first lower than it did 

 originally, but exhibits a tendency to recover. The most 

 characteristic features were discussed thirty years ago by Dr. 

 Balfour Stewart. They have also been the subject of a pamphlet 

 liy Mr. Edward Whymper, who gives the results of a number 

 lit interesting long period experiments. 



The present paper treats of how the differences between the 



readings with pressure descending and ascending in a normal 



jnessure cycle, such as the Kew test, varies throughout the 



range, and how the sum of these differences varies from one 



range to another. It investigates how the error, as pressure is 



reduced, varies w'th the rate of fall of pressure (when uniform), 



liow the fall of reading at a low stationary pressure increases 



with the time, depends on the pressure, and varies with the 



rate of the previous fall of pressure, and how the recovery after 



a pressure cycle progresses with the time, and is modified by 



he nature of the previous pressure changes. The influence of 



■ubsidiary stoppages is investigated, and experiments are 



-cussed showing the influence of temperature. 



The opportunity is taken of considering the secular change 



zero, and also changes in the elastic and the after-effect 



-perties. 



Algebraic and exponential formulae are obtained for such 

 cnomena as the variation of the differences of the descending 

 1 ascending readings throughout a pressure cycle, the de- 

 iidence of the sum of such differences on the range, the fall of 

 iding at the lowest pressure and the final recovery. A theory, 

 some extent empirical, is built up, leading to mathematical 

 ults, depending on only three arbitrary constants, for the 

 cliaviour of an aneroid in the ordinary Kew test over any 



range. 



The large differences brought to light between different 

 leroids show that the means of markedly raising the average 

 L already at the makers' disposal. The present inquiry shows 

 ( arly how the effects of tentative improvements may be 

 certained. 



Physical Society, June 24.— Mr. Walter Bailyin the chair. 



I— I'rof. Carus- Wilson exhibited an apparatus to illustrate the 

 action of two electric-motors coupled in such a way as to admit 

 of their rotating at different speeds. The two shafts are placed 

 in line, and each is fitted with a bevel-wheel, gearing into an 

 intermediate wheel. The axis of the intermediate wheel is at 

 ; light angles to the line of the motor-shafts, and is free to rotate 



NO. 1496, VOL. 58] 



in a plane at right angles to that line. The motors can be made 

 to rotate at different speeds by altering the strength of the 

 magnets of either or both. The motion of the intermediate 

 wheel depends upon the difference of the two speeds, or upon 

 their mean, according to their relative directions of rotation. A 

 simple graphic construction enables the action to be predeter- 

 mined for any given load on the intermediate wheel. Calling 

 the two motors A and B, and the intermediate wheel C, lines 

 can be drawn on a base of current to represent the speeds and 

 the torques for each motor. If the motions of A and B are in 

 the same direction, the load or torque is the same on each, and 

 of similar sign. Hence, as the load on the wheel C is increased, 

 the speeds of A and B tend to become equal (if A had been 

 running faster than B) ; and for a certain load on C the speeds 

 of A and B will be equal. If the load on C is further increased, 

 B will run faster than A. Also, there will be a certain value 

 for the load on C at which the motion of A will reverse. A 

 further increase of the load on C will bring C to rest, A and 

 B then rotating at equal speeds in opposite directions. When 

 the load on C is nothing, let the motors rotate in opposite direc- 

 tions, A running faster than B. The motion of C now depends 

 upon the difference of speeds of A and B. When a load is put 

 on C, the motion of A is retarded, while that of B is assisted, 

 hence B takes less current, and A takes more. The torques on 

 the two motors, due to the load on C, are now of equal amount, 

 but of opposite sign. As the load on C is increased, the speed 

 of A is reduced, and that of B increased, until the two are 

 equal, and C comes to rest. B is now acting as a generator, 

 and sending current into A. If the load on C is simply that 

 due to friction, the process cannot be carried further. But if the 

 load on C is reversed, the speed of B becomes greater than that 

 of A, and the motion of C is reversed. In the steering gear 

 designed by the Union Electricitats Gesellschaft, the inter- 

 mediate wheel is made to actuate a rudder by differential 

 action. The motion is reversed by making the speed of 

 one motor greater or less than that of the other. — 

 Mr. Quick then exhibited Weedon's apparatus for the measure- 

 ment of the expansion of solids. This method is claimed to 

 be independent of knowledge of optics on the part of the 

 student. The expansion is read directly by means of a pair of 

 micrometers. Precautions are taken to prevent errors due to 

 radiation. Mr. Lehfeldt asked what precautions were taken 

 to prevent the movement of the micrometer supports. Mr. 

 Stansfield described a form of apparatus in use at Chelsea 

 Polytechnic ; it was a simple contrivance, in which changes of 

 length were measured by a micrometer. Mr. Quick, replying, 

 thought the instrument referred to by Mr. Stansfield pre-supposed 

 a knowledge of optics. — Mr. Lehfeldt then read a paper by Dr. 

 Donnan on the theory of the Hall effect in a binary electrolyte. 

 In 1883 Roiti investigated the subject of a possible Hall effect 

 I in electrolytic solutions. He failed to obtain any positive result. 

 j Recently the question has been examined by Bagard, who 

 i noticed certain effects in aqueous solutions of zincic and cupric 

 I sulphates. Meanwhile, negative results have been observed by 

 ■ Florio. The author therefore discusses what effect might be 

 [ expected by theory, on somewhat the same lines as those of Van 

 I Everdingen, jun., taking a more general case. So far as the 

 present discussion goes, the author's theory is wholly in favour 

 of the negative results of Roiti and Florio. It would appear 

 that Bagard measured a phenomenon not contemplated by the 

 theory as stated in the present treatment. Van Everdingen 

 originally supported the positive results of Bagard ; but his 

 work, unfortunately, was rendered incorrect by the accidental 

 omission of a numerical factor. He has since discovered the 

 slip in his calculations, and now agrees with the author's con- 

 clusions. — The Chairman proposed votes of thanks to the authors, 

 and the meeting adjourned until October, this being the last of 

 the session. 



Linnean Society, June 16. — Dr. A. GUnther, F.R.S., 

 President, in the chair. — Prof. J. B. Farmer and Mr. W. G. 

 Freeman demonstrated the action of germinating peas, cress, 

 and barley in causing the deoxidation of a watery solution of 

 methylene blue to a colourless liquid on shaking up the latter 

 with air, while on adding a drop of hydrogen peroxide the blue 

 colour was restored. Green plants placed in the solution were 

 found to act in a manner precisely similar to the seedlings, 

 though the action may be modified by assimilation in sufficient 

 light (see p. 185). — Mr. F. Enock exhibited and made 

 remarks on the eggs of an hemipterous insect containing living 

 parasites (/V«/w/V/4/a aquatica), of whose life-history and habits 



