June 15, 1893] 



NA TURE 



155 



owing to the continuance of easterly winds, many of the maxi- 

 mum readings have been below 60°. During the early part of 

 the present week the anticyclone decreased in intensity and 

 began to move slowly eastward, while a depression which lay 

 over the south-west of France moved northward, causing the 

 barometer to fall generally over our islands, and on Tuesday 

 night, the 13th inst., a thunderstorm occurred at Jersey, while a 

 further rise of temperature occurred over the southern portion 

 of the kingdom, the maximum at Cambridge reaching 81°. 

 The Weekly Weather Report of the loth inst. showed that 

 the greatest excess of temperature occurred in Ireland and 

 the Channel Islands, where it was 5° above the mean. The 

 rainfall just equalled the mean in the east of Scotland and 

 the north of Ireland only. Bright sunshine was above the 

 usual amount everywhere ; the percentage of duration ranged 

 from 29 to 38 in Scotland, from 36 to 40 in Ireland, and 

 from 40 to 67 over England, while in the Channel Islands the 

 percentage was as high as 75 of the possible amount. 



We are indebted to Dr. A. Buchan for the discussion and 

 publication (in vol. ix. of the Journal of the Scottish Meteoro- 

 logical Society) of a very valuible series of mean monthly and 

 yearly temperatures for London and vicinity for 130 years, from 

 1763 to 1S92. The only interruptions of the continuity of this 

 long series occurred in May 1777 and July 1780, and the means 

 for these months have been interpolated. For the 130 years the 

 mean temperature of London is 50° '2. The highest mean tem- 

 perature of any month was 74°! in July 1783, and the lowest 

 25° '9 in January 1795, the difference being 48° '2. The warmest 

 seasons were, winter, 1779, 8° -3 above the normal value ; 

 spiing, 1811, + S°'2; summer, 1783, -hS'^; and autumn, 

 1777, -h5°i. The coldest seasons were, winter, 1814, 6'''5 below 

 the normal ; spring, 1837, - 6° 7 ; summer, 1816, - 4'6; and 

 autumn, 1877, -4'"I. The year 1783 had the highest mean annual 

 temperature, being 5° '2 above the normal for the year ; and 1816, 

 the lowest, being 3°'5 under the normal. Dr. Buchan states 

 that much labour has been spent in searching for evidence of 

 cycles, bat that it cannot be said that the results show more 

 than highly interesting resemblances and contrasts among the 

 months, and that in whatever way the periods are viewed, they 

 suggest no appearance of a cycle. But a tendency is shown of 

 types of high and low temperature to prolong themselves during 

 months, seasons, and years. Following this paper is an equally 

 important discussion by the same author of the temperature of 

 the north-east of Scotland for 129 years, from 1764 to 1892 

 from observations taken at Gordon Castle and other places. 



At the instance of Herr von Helmholtz, and with the support 

 of the Berlin .\cademy of Sciences, Drs. Franz Richarz and 

 Otto Krigar-Menzel have undertaken a remarkable series of 

 experiments for the purpose of determining by weighing the 

 diminution of gravity as we ascend from the surface of the 

 earth. The method was theoretically the following : — To each 

 pan of an ordinary balance is attached another pan by means 

 of a rod about 2 m. long. Two sensibly equal masses are 

 placed in the left upper and the right lower pan respectively. 

 The gravitational attraction being stronger on the latter weight, 

 a difference will be indicated by the balance. On removing 

 the left weight from the upper to the lower pan, and the tight 

 weight from the lower to the upper, the difference acts in the 

 opposite direction, and half the mean of the two differences 

 gives the decrease of gravity with the height. It is almost 

 needless to say that the experiment was one of very great 

 delicacy and difficulty. It was performed in an earth-covered 

 casemate of the citadel of Spandau, partly in order to utilise 

 a mass of lead weighing about a hundred tons to determine the 

 attraction exerted by it. The necessary preparations were 

 begun in 1887, and the main part of the observations has only 

 NO. 1233, VOL. 48] 



just been concluded. The difference between the values of^'-at 

 two points, one 2'26 m. above the other, was found to be 

 6'523 X I0"°. The calculated value was 6 '970 x 10'. The 

 difference may be due to a density of the strata below the- 

 station being less than the average. 



A THERMOSTAT for the comparison of standard thermometers 

 between the temperatures of 50^ and 300' C. is described by 

 Herr A. Mahlke in the current number of the Zeitschrift Jiir 

 Instrumentenkunde. It is used at the Physikalisch-Technische 

 Reichsanstalt for the purpose of maintaining the thermometers 

 to be compared at certain temperatures for which the boiling 

 point of some substance is not available. It consists essentially 

 of an oil-bath in a copper cylinder surrounded by another 

 copper cylinder, the space between the two being filled with 

 air. Heat is applied to the outer vessel. The heated air 

 warms the inner vessel by circulating round it, there being a 

 clearance of 2 or 3 cm. all round. Special precautions are 

 taken to keep the level of the oil in the inner vessel constant, 

 and the temperature of the oil uniform throughout. Both 

 cylinders are closed with lids containing holes through which to 

 insert the thermometers. The oil is kept circulating by means 

 of two propellers enclosed in vertical copper cylinders open at 

 both ends. Their axes project through the outer lid, and are 

 provided with pulleys rotated by means of a small water motor. 

 The whole arrangement is designed to keep the entire body of 

 oil in motion, so as to prevent unequal heating. Surplus oil, 

 due to expansion, and any oil-vapour that may be evolved, are 

 drawn off through a siphon leading through the walls of the 

 cylinders into a refrigerator. The apparatus has worked veiy 

 well, the variations of temperalure not exceeding the average 

 errors in reading the thermometer scales. 



The extremely high cost of high resistances made of 

 metallic wire causes the discovery of a cheap substitute 

 to be a matter of considerable importance. Most of 

 the substitutes hitherto proposed, such as liquid resist- 

 ances, pencil marks on glass or ebonite, &c., are subject 

 to the objection that they have an extremely high tem- 

 perature coefficient, and in the case of the pencil mark, on 

 account of the extreme thinness of the onducting material, the 

 rubbing off of a few particles causes a great increase of resist- 

 ance. These resistances also depend in a considerable degree 

 on the electromotive force to which they are subjected. The 

 Elettricisla for May contains the description of a new material 

 for the construction of high resistances, discovered by E. Jona 

 which is said to be free from most of these defects. He uses an. 

 ebonite tube which is filled with a mixture of graphite and un- 

 vulcanised ebonite in suitable proportions. The mixture is then 

 vulcanised, when it hardens and adheres to the containing tube. 

 Metal cups fitted with binding screws are fixed to the ends. In 

 this way a resistance of a megohm can easily be obtained in a 

 tube 10 cm. long and of 15 mm. diameter. 



In a recently-published number of the proceedings of the 

 Cambridge Philosophical Society there is an interesting paper 

 by Messrs. Griffiths and Clark on the determination of low 

 temperatures by means of platinum thermometers. Acting on. 

 the suggestion of Profs. Dewar and Fleming, that from obser- 

 vations on the resistance of certain pure melals (including, 

 platinum) at very low temperatures, it would appear as if the 

 resistance vanished at absolute zero, the authors have calculated 

 by means of Callendar and Griffuhs' method the temperature 

 at which the resistance of several platinum thermometers, whose 

 accuracy had been severely tested, would be zero. The values- 

 obtained seemed to corroborate the conclusions arrived at by 

 Profs. Dewar and Fleming as the mean value found for the 

 temperature at which the resistance would be zero is -273'"96. 

 This gives a convenient method of graduating a platinum ther" 



