S02 



NATURE 



{March 21, 1889 



in direction to the resting difference, some apparently single and 

 others multiple. These lasted from twenty to thirty seconds and 

 suddenly ceased. 



We have repeated this observation thirty or forty times, and 

 feel ourselves justified in concluding that we have obtained 

 evidence that during a cortical epileptiform discharge the electro- 

 motive changes in the spinal cord are exactly parallel as regards 

 the character of their sequence to the convulsions of the muscles 

 as recorded by the graphic method. It remains to be stated that 

 after removal of the cortex we have obtained an effect in the 

 electrometer when the corona radiata was stimulated. This 

 effect was only present during the period of excitation, no 

 rhythmical after effect ever being observed. Its character was 

 prolonged, and resembled the first persistent stage referred to 

 above. 



In conclusion, we consider that, since by the method we have 

 adopted the influence of the lumbar bulbo-spinal centres is 

 excluded, the existence of the epileptic rhythm in the dorsal 

 regions of the spinal cord points to its being almost entirely of 

 cortical origin. 



Physical Society, February 23. — Prof. Reinold, President, in 

 the chair.- — Dr. J. W. Waghome read a note on the measurement 

 of electrical resistance, showing that two resistances may be 

 compared by joining them in series with a battery, and observ- 

 ing the deflections of a galvanometer connected successively with 

 their terminals. The resistances are proportional to the currents 

 which pass through the galvanometer in the two cases, provided 

 they are large compared with that of the battery, or are not very 

 different from each other. By using a rocking key, the method 

 is rendered expeditious, and the galvanometer resistance need 

 not be known. — On a new polarimeter, by Prof. S. P. Thomp- 

 son. The author gave a risuvic of the ordinary methods of de- 

 termining the position of the plane of polarization, pointing out 

 their advantages and disadvantages ; and exhibited his new 

 polarimeter, in which two black glass mirrors, placed at a 

 small angle (about 2^°), are used to polarize the light in 

 two different planes. By using a modified Nicol as analyzer, 

 the plane can be determined to one-tenth of a degree, when 

 the substance examined does not absorb much light ; but, 

 for dark-coloured liquids, the author prefers to use one of his 

 " twin prisms," described before the British Association in 1887, 

 as polarizer, in which the planes of polarization are 90° apart. 

 A method of dividing a polarized beam into two parts inclined 

 at a variable angle, by means of a combination of quarter-wave 

 plates of mica, was described, in which the two halves of the 

 field are similarly coloured. Mr. Glazebrook considered Poynt- 

 ing's glass cell, with different thicknesses of active solution, a 

 very convenient means of obtaining two beams polarized at a 

 small angle, as, by altering the strength of the solution, the 

 angle may be varied at will. — Prof. Thompson also read a note 

 on the formation of a cross in certain crystal structures. Several 

 specimens (including benzoic acid, stalactite, Eno's salt, &c.), 

 which exhibit a radial structure, and show a cross when exa- 

 mined by polarized light, were thrown upon the screen, and the 

 fact that the cross remains stationary when the specimens are 

 rotated demonstrated. Similar effects were produced by mica 

 sectors arranged radially, thus showing the stationary cross to 

 be caused by the light not being analyzed in those directions. — 

 On electrical measurement, by Prof W. E. Ayrton, F. R. S., 

 and Prof. J. Perry, F.R.S. In a paper on winding voltmeters, 

 read before the Society in 1885, the authors showed, on the 

 assumption that the thickness of insulating covering on wires 

 was proportional to their diameter, that instruments wound with 

 copper wire gave a less heating error than similar ones wound 

 with German silver. Since then, platinoid has been introduced, 

 and the electrical constants of phosphor-bronze determined. 

 Further, a remarkably simple relation between the volts corre- 

 sponding to a given deflection on a given type of instrument, 

 and the resistance per unit length of the wire used in winding 

 it, has been suggested by Mr. Crawley. Suppose F = num- 

 ber of ampere turns required to produce the deflection P, and U 

 the half area of section and volume of the coil respectively, and 

 d and D the diameters of the bare and covered wires, then — 



U . 



(pay) where K is a constant depending on the type of instrument. 

 Since '^'',, is the resistance per unit length, t/ie volts required to- 



produce a given deflection are proportional to the resistance per 

 unit length of the wire used, 'whatever he the material of the 

 wire or thickness of the insulation. Taking this into account, 

 and using a more accurate value for the thickness of the covering, 

 it is shown that the four metals above referred to arrange them- 

 selves in the following order of merit when used for high- 

 reading voltmeters — platinoid, phosphor bronze, german silver, 

 and copper ; and for comparatively low-reading instruments, the 

 last two change places. As a standard ammeter of great range, 

 a circuit containing a Depretz D'Arsonval galvanometer is 

 shunted by a wide sheet of thin platinoid, and by altering the 

 resistance in the galvanometer circuit, the sensibility may be 

 varied in known proportions. An instrument on this plan has 

 been arranged to measure any current from OT to 800 amperes 

 to one-quarter per cent., and the same galvanometer in series with 

 various resistance coils is used as a standard voltmeter of 

 practically unlimited range. Whilst arranging these standards 

 it has been found that the deflections are not generally propor- 

 tional to the currents, and the discrepancy traced to the centre 

 of gravity of the swinging coil not being in the line of suspen- 

 sion. By replacing the bottom torsion wire by a long thin 

 spring the defect may be remedied. As relating to calibration 

 curves of instruments, it was mentioned that in '• Siemens's 

 dynamometer" the "square law " is not correct, probably owing 

 to distortion of the spring. Referring to "hot wire volt- 

 meters," in which the sag of a wire heated by the current is 

 measured by a magnifying spring, the authors remark that, in 

 their original paper on the subject, they neglected the change of 

 length due to change of stress in the wire, and subsequently 

 their assistant, Mr. Bourne, found that maximum sensibility 

 was never co-existent with minimum sag. The sag which gives 

 maximum sensibility depends on the initial stress in the 

 wire, and by altering the initial sag the instruments may be 

 compensated for changes of temperature of the room.. In the 

 present paper the mathematical treatment is more rigorous, and 

 the results are in accordance with experiment. A voltmeter 

 intended for use with "electric welders," which deflects 

 about 300° for 2 volts (direct or alternating), and is graduated to 

 i/ioo of a volt, was exhibited, and used to measure the resist- 

 ance of a storage cell. Dr. Thompson suggested that the want 

 of proportionality of D'Arsonval galvanometers might be due to 

 lateral displacement of the coil caused by the current in the 

 torsion wires crossing a magnetic field, but from experiments 

 with pivoted coils the authors thought this improbable. — Prof. 

 Riicker read a note on the dimensions of electro-magnetic units, 

 by Prof G. F. Fitzgerald, F.R.S., which suggests that specific 

 inductive capacity and permeability be assumed to be of dimen- 

 sions - (slowness) ; if this be done the dimensions of quanti- 

 ties expressed in electrostatic and electro-magnetic measure 

 become identical. The author also states that it seems most 

 likely that inductive capacities are related to the reciprocal 

 of the square root of the mean energy of turbulence of the eiher. 

 Prof. Riicker remarked that in his recent paper on the subject 

 he considered it important to retain K and jx as secondary funda 

 mental units, and Mr. Blakesley did not concur with Prof. 

 Fitzgerald's suggestion. — A photograph of crystal models, by 

 Mr. R. T. Anderson, of Belfast, was exhibited at the meeting. 



Chemical Society, February 21.— Mr. W. Crookes, F.R.S. , 

 in the chair. — The following papers were read : — Note on the 

 decomposition of potassic chlorate by heat in the presence of 

 manganic peroxide, by Prof H. McLeod, F.R.S. The author 

 concludes from his experiments that the reaction which ensues 

 when potassic chlorate is heated with manganic peroxide most 

 probably consists in the formation of potassium permanganate 

 chlorine, and oxygen in the first instance ; that the permanganate 

 as rapidly as it is produced is decomposed by the heat into 

 potassic manganate, manganic peroxide, and oxygen ; and that 

 the resulting potassic manganate is acted on by chlorine generated 

 by the action of the peroxide on some fresh chlorate, forming 

 potassic chloride, manganic peroxide, and oxygen, so that the 

 peroxide is being continually reproduced. The quantity of 

 chlorine evolved corresponds to only a very small proportion of 

 the manganic peroxide present, so, if the first action really takes 

 place, the chlorine must be absorbed and employed in converting 

 the potassium into chloride. — The vapour-density of hydrogen 



