September 4, 1919] 



NATURE 



view of this invitation there was to have been a 

 preliminary meeting of the directors of some of the 

 laboratories concerned in Berlin in September, 1914; 

 much correspondence had passed and all arrange- 

 ments were complete. Representatives of the Reichs- 

 anstalt visited the National Physical Laboratory in 

 June, 1914, to compare standards. Scd dis aliter 

 visitiit. So far as England and America are con- 

 cerned, uniformity has been secured over the range 

 from -182° C. to about 1100° C. by agreement 

 between the directors of the Bureau of Standards and 

 the National Physical Laboratory, and a common 

 scale is in use at those institutions. 



Another important question dealt with by the con- 

 ference related to the standardisation of end measures 

 of length. M. Perard presented an interesting report 

 on the results of his investigation as to the methods 

 of comparing end measures and line measures, and 

 particularly as to the accuracy with which a length 

 gauge built up of a number of Johannsen slips, the 

 use of which has become so familiar during the war, 

 represented the sum of the lengths of the individual 

 gauges. He states as the result of his measurements 

 that it is necessary to add an amount of about eight- 

 hundredths of a micron for each contact in the com- 

 plex gauge, a quantity which is negligible for all 

 practical purposes when using the standards. But 

 the report led to an interesting discussion on another 

 point. The metre is standard at 0° C. ; end gauges 

 are used chiefly in engineering work, and the tem- 

 perature in engineering shops is probably between 

 15° C. and 20° C. If, then, a gauge is standard at 

 zero, it is necessary when using it at 20° to know 

 its temperature coefficient with some accuracy ; two 

 gauges, one of steel, ithe other of gun-metal (say), 

 both correct at the standard temperature, differ 

 appreciably at the temperature of use. Accordingly 

 it has been proposed to standardise gauges for indus- 

 trial purposes at a temperature more nearly that of 

 an average shop ; the correction which would then 

 be required would in most cases be so small as to 

 be negligible. English practice, of course, secures 

 this; our measures are standard at 62° F. {16'] ° C). 

 In .\merica a temperature of 20° C. has been 

 suggested. 



The International Committee of Weights and 

 Measures had in igoq preferred to accept 0° C. as 

 the standard temperature, and the French Technical 

 Section of Artillery had adopted that in the standards 

 employed by it. After discussion the matter was 

 referred to a soecial committee consisting of Dr. 

 Foerster and MM. Blumbach,' Egoroff, Guillaume, 

 and Perard, and on their recommendation the con- 

 ference decided to invite the International Committee 

 to pursue its investigations with the view of reaching 

 as complete knowledge as possible of the metrological 

 properties of steels, as well as of the construction, 

 standardisation, and method of use of end standards 

 of length. Sir David Gill pointed out that, while 

 recognising that the adoption of 0° C. as the standard 

 temperature was theoretically desirable, he accepted 

 the resolution as one which reserved a final decision 

 on the question until the completion of the experi- 

 ments. The question is a vital one if international 

 standardisation is to become a reality. Its import- 

 ance was illustrated during the war. A number of 

 screw gauges about 2 in. in diameter were ordered 

 on the Continent; the first batch received failed to 

 pass inspection. They had been made correct at the 

 freezing-point, and the expansion between that tem- 

 perature and the English standard, 62° F., brought 

 them outside the limits for acceptance. A visit t» 

 the Continental works was necessary to explain the 



» M. Blumbach's many fncnd.s will be elad to know that he is well : .t 

 messace ha^ h'en received from him a<ikin£; for help to reconstitute the 

 Weights and Me-^siires Deparrment in Rn«sia. 



NO. 2601. VOL. ioa.1 



point and put matters straight, causing delay and 

 loss when time was of great importance. 



A third decision of the conference is of special 

 interest to meteorological circles. In correcting a 

 barometer to a standard, sea-level at latitude 45° has 

 been accepted, and a formula due to Clairaut, but 

 modified by Broch, based on the accepted value of g 

 in latitude 45° and its variations with latitude, has 

 been employed. The accepted value for g at sea- 

 level at latitude 45" has been 980-665 cm. /sec.'; 

 recent observations, however, lead to 980615 cm./sec.^ 

 as the more correct value,^ and it became a question 

 whether to modify the normal standard or not- The 

 conference decided against any modification in view 

 of the fact that the figure 980665 cm. /sec' had 

 received legislative .sanction in many countries. They 

 agreed, however, that observations into which the 

 local value of g entered should be reduced, not by the 

 usual formula, but by means of a numerical factor 

 "determined, if possible, directly for the locality in 

 question." 



Now the unit of barometric pressure is the millibar 

 — a pressure of 1000 dynes per sq. cm. — and the 

 length of the column of mercury the weight of 

 which gives rise to a pressure of one millibar is 

 known as a baromil. A barometer the scale of which 

 is graduated in baromils reads pressure in millibars. 

 The length of the baromil is dependent on the value 

 of g, and so varies with the locality,', but it has been 

 agreed by the International Conference of Meteoro- 

 logists to take as standard the value of g at sea-level 

 in latitude 45"', and the relation of the baromil to 

 the mm. or the inch has been evaluated on the hypo- 

 thesis that the standard value of g is 980617 cm. /sec.', 

 the value assigned to it by Helmert. The decision, 

 then, of the International Conference of Weights and 

 Measures to retain as the standard value of g the 

 old figure 980665 involves, if it be used, a reduction 

 of about 5 parts in 100,000 in the height of the 

 column of mercury producing under standard condi- 

 tions a pressure of one millibar. The certificates 

 usually issued with barometers state the temperature 

 at which the scales read pressure in millibars under 

 standard conditions. If reduction is to be made to a 

 nominal standard of gravity of 980665 cm., these 

 temperatures will all need to be reduced correspond- 

 inglv. The difficulty, of course, is avoided if we take 

 the true value of g at latitude 45° instead of the 

 standard value as the figure to be used in calculating 

 the length of the baromil, i.e. if we assume that this 

 is one of the cases in which a numerical factor 

 directly determined is to be employed, and some such 

 course as this no doubt will be adopted. 



A fourth matter of some interest which was dis- 

 cussed was the proposal to adopt in future legislation 

 for metric countries the M.K.S. system of units, in 

 which the unit of length is the metre, that of mass 

 the kilogram, and of time the second. On this system 

 the unit of force is the "Newton," the force required 

 to produce per second in a mass of i kilogram a 

 velocity of i metre per second. Thus, 



I Newton = 1000 X 100 C.G.S. units=io^ dynes. 



The unit of work will be the work done by a Newton 

 in moving its point of application i metre or 

 lo'xio" ergs, and this is the joule of the C.G.S. 

 system. Accordinglv, the unit of power is the watt. 

 ' As a result of the discussion the International Com- 

 mittee was invited to continue the study of all ques- 

 tions connected with legislation based on metrical 

 units. 



.At the last session of the conference M. Battistella, 

 the Italian delegate, raised a far-reaching question. 

 The work of the conference and of the committee 



•^ The value given by Helmert is 98o"6i7 cm. /.sec.-. 



