April 26, 1906J 



NA TURE 



623 



tically finished i/40th of a second after it begins, and 

 that i/30th of a second after the attainment of maximum 

 pressure the gas in the vessel may be regarded as a 

 mixture of CO,, steam, and inert gases in chemical 

 equilibrium. 



The pressure of the ignited gas at the centre of the 

 vessel is increased during the spread of the flame from 

 one atmosphere to six. During this time it loses no heat, 

 and the rise of temperature observed is from 1200 C. to 

 1900° C. It follows that between these limits of tempera- 

 ture the average value of 7 for these gases is 1 25. 



With a weaker mixture containing one volume of gas 

 and twelve of air the spread of the flame is very much 

 slower, about i\ seconds elapsing before all the gas is 

 burned. Owing to the slow propagation of the flame, con- 

 vection currents play an important part during the process 

 of ignition ; the burned gases rise to the top of the vessel, 

 and the last portion of gas to be ignited is not close to the 

 wall, but immediately under the spark, and a short distance 

 from it ; but though the flame is propagated very slowly, 

 the combustion of any given portion of gas, when once 

 started, proceeds almost as rapidly . as in the stronger 

 mixture. There is no " after-burning " in the sense of 

 the slow completion of a reaction already begun. Within 

 i/ioth of a second before the time of maximum pressure 

 some gas is still unburnt ; within i/ioth of a second after 

 all the gas is completely burned, and the mixture every- 

 where in chemical equilibrium. 



Incidentally, the difference of temperature between a 

 fine wire immersed in the gas and the temperature of the 

 gas is determined by comparing the temperatures of two 

 wires, one having double the diameter of the other, placed 

 close together in the same explosion. The error due to 

 radiation is thus found, and it is shown that if a wire 

 1 /500th of an inch in diameter is getting hotter at the 

 rate of 1300 C. per second, then it must be 200° C. colder 

 than the gas surrounding it. The results are used to 

 find the actual temperature of the gas from that of a wire 

 i/ioooth of an inch diameter immersed in it, and the 

 conclusion is drawn that the temperatures in a gas-engine 

 cylinder cannot be obtained by the use of a wire thicker 

 than this, except by applying corrections amounting to 

 several hundred degrees centigrade. 



The bearing of the results on the question of " after- 

 burning " in the gas engine is discussed, and it is shown 

 that the high specific heat of the products of combustion, 

 together with some loss of heat during the passage of the 

 flame through the compression space, accounts for all the 

 peculiarities of the gas-engine diagram. The form of 

 diagram obtained with weak mixtures is due simply to 

 the very slow propagation of the .flame, and not to any 

 delay in the attainment of chemical equilibrium at a point 

 which the flame has already reached. 



March 15. — " A Discussion of Atmospheric Electric 

 Potential Results at Kew, from selected Days during the 

 Seven Years 1898 to 1904." By Dr. C. Chree, F.R.S. 



The paper contains an analysis of atmospheric electricity 

 results at Kew on selected fine-weather days — usually ten 

 a month — from 1898 to 1904. 



All days were excluded when rain fell or negative 

 potential was recorded. All data are given in absolute 

 measure (volts per metre). The diurnal inequalities for 

 individual months and the year are represented by curves. 

 These all show two distinct daily maxima and minima. 

 The minima always occur near 4 a.m. and 2 p.m. The 

 times of the maxima are more variable, the day interval 

 between the two being longer in summer than in winter. 



The highest mean potential gradient occurs in December. 

 Whilst the amplitude of diurnal inequality is greatest in 

 mid-winter, the ratio in which it stands to the mean daily 

 value is then least. The diurnal inequalities for the 

 several months are analysed in 4-wave Fourier series. 

 The 12-hour term is, in general, the most important ; the 

 changes in its amplitude and phase angle throughout the 

 year are comparatively small. The 24-hour term is much 

 larger in the winter than in the summer months, and its 

 phase angle varies greatly. Attention is also given to the 

 phenomena of individual days. The difference "between the 

 highest and lowest hourly values averages two and a half 

 times the amplitude of the regular diurnal inequality, and 

 is fully larger than the mean value for the day. 

 NO. I904, VOL. 73] 



Of various meteorological elements temperature is found 

 to have much the most marked influence, high mean 

 potential and large diurnal range of potential being 

 associated with low temperature in every month of the 

 year, except the hottest (July). 



An appendix compares the diurnal inequalities of 

 potential and barometric pressure. Diurnal inequalities 

 were got out for each month of the year for the baro- 

 metric pressure at Kew for an 11-year period. The 

 similarity between the diurnal inequalities of the two 

 elements is found to be confined to the 12-hour terms; 

 the 24-hour terms present diametrically opposed phenomena 

 in the two cases. The afternoon minimum and evening 

 maximum of potential are in every month notably in 

 advance of those of barometric pressure. If any relation- 

 ship of cause and effect exists between the regular diurnal 

 changes in the two elements, the pressure change would 

 seem to be the effect, the potential change the cause. 



Geological Society, April 4. — Mr. R S. Herries, vice- 

 president, in the chair. — A case of unconformity and 

 thrust in the Coal-measures of Northumberland : Prof. 

 G. A. L. Lebour and Dr. J. A. Smythe. The sections 

 described occur on the coast north of the Tyne, near 

 Whitley Sands, between Table Rocks and Briar-Dene Burn. 

 The base of the " Table-Rocks Sandstone " is found to 

 rest unconformably upon a series of alternating shales and 

 sandstones, among which is a well-marked band of clay- 

 ironstone crowded with Carbonicola acuta, one of those 

 " mussel-bands " which are found to be perhaps the most 

 remarkably persistent strata in the north of England 

 Carboniferous rocks. — The Carboniferous succession below 

 the Coal-measures in North Shropshire, Denbighshire, and 

 Flintshire: Dr. YVheelton Hind and J. T. Stobbs. This 

 paper opens with a critical account of previous research 

 among the Carboniferous rocks of North Wales. Then 

 follows a detailed account of the various beds, exposed in 

 numerous quarries worked for road-metal, iron manufac- 

 ture, lime, cement, chert, or building-stone. Fossil lists 

 are given from each exposure of importance. A range table 

 is given of the chief brachiopods and corals, and the 

 palaeontological sequence is compared with that occurring 

 at Bristol and in the north of England. 



Chemical Society, Aoril 5. — Prof. R. Meldola, F.R.S., 

 president, in the chair. — An improved apparatus for measur- 

 ing magnetic rotations and obtaining a powerful sodium 

 light : W. H. Perkin, sen. The improved apparatus 

 consists of a short but very powerful coil carrying a 

 powerful electric current. The coil is cased with steel, and 

 has a 3-inch gun-metal tube through the centre, the interior 

 of this being the position of the magnetic field. The glass 

 measuring tubes are supported in this tube in a metal 

 trough which can be kept at any required temperature. 

 A method of obtaining a powerful sodium light was de- 

 scribed, which consists in heating a platinum boat contain- 

 ing sodium chloride by a small oxygen-coal gas flame. 

 This causes the sodium chloride to volatilise, and the 

 vapour, passing into a flame produced by a large Bunsen 

 burner, gives a very intense, yellow light, which can be 

 maintained for a long time. — The rusting of iron : G. T. 

 Moody. The explanation of rusting as a process involving 

 the production of hydrogen peroxide, as advanced by 

 Dunstan, is directly negatived by experimental evidence, 

 which shows that atmospheric corrosion results first from 

 the interaction of iron and carbonic acid, whereby ferrous 

 salt is formed, and subsequently from the more or less 

 complete oxidation of ferrous salt by oxygen. It is found, 

 moreover, that the composition of iron rust is not fairly 

 represented by the formula Fe 2 0.,(OH) 2 , as stated by the 

 foregoing investigator. — The estimation of carbon in soils : 

 A. 5. Hall, N. H. J. Miller, and N. Marmu. The soil 

 is treated with chromic acid, and the resulting gases passed 

 over a short length of copper oxide. The carbon dioxide 

 formed is absorbed by alkali and estimated by double 

 titration. — Electrolysis of salts of ftS'-dimethvlglutaric 

 acid : J. Walker and J. K. Wood. — Bromo- and hydroxy- 

 derivatives of /9|8/3'0'-tetramethyIsuberic acid : J. K. Wood. 

 — Some new o-xylene derivatives : G. Stallard. — A new 

 solvent for gold, preliminary note : J. Moir. The author 

 finds that gold-leaf dissolves fairly readily when floated on 

 an acid solution of ordinarv thiocarbamide, and solu'ion 



