August 16, 1907] 



SCIENCE 



195 



of metal has been reproduced with more 

 precision, and is known with higher ac- 

 curacy in terms of many secondary stand- 

 ards, than is the length of any other 

 standard in the world; but it is, after all, 

 liable to destruction and to possible secular 

 change of length. For these reasons it can 

 not be scientifically described otherwise 

 than as a piece of metal whose length at 0° 

 C. at the epoch a.d. 1906 is = 1,553,164 

 times the wave-length of the red line of the 

 spectrum of _ cadmium when the latter is 

 observed in dry air at the temperature of 

 15° C. of the normal hydrogen-scale at a 

 pressure of 760 mm. of merciiry at 0° C. 



This determination, recently made by 

 methods based on the interference of light- 

 waves and carried out by MM. Perot and 

 Fabry at the International Bureau of 

 Weights and Measures, constitutes a real 

 advance in scientific metrology. The re- 

 sult appears to be reliable within one ten- 

 millionth part of the meter. 



The length of the meter, in terms of the 

 wave-length of the red line in the spectrum 

 of cadmium, had been determined in 1892 

 by Michelson's method, with a mean result 

 in almost exact accordance with that just 

 quoted for the comparisons of 1906; but 

 this agreement (within one part in ten 

 millions) is due in some degree to chance, 

 as the uncertainty of the earlier determina- 

 tion was probably ten times greater than 

 the diiference between the two independent 

 results of 1892 and 1906. 



We owe to M. Guillaume, of the same 

 International Bureau, the discovery of the 

 remarkable properties of the alloys of 

 nickel and steel, and from the point of 

 view of exact measurement the specially 

 valuable discovery of the properties of that 

 alloy which we now call " invar." He has 

 developed methods for treatment of wires 

 made from this alloy which render more 

 permanent the aiTangement of their con- 



stituent molecules. Thus these wires, with 

 their attached scales, may, for considerable 

 periods of time and under circumstances of 

 careful treatment, be regarded as nearly 

 invariable standards. With proper pre- 

 cautions, we have found at the Cape of 

 Good Hope that these wires can be used for 

 the measurement of base lines of the 

 highest geodetic precision with all the ac- 

 curacy attainable by the older and most 

 costly forms of apparatus ; whilst with the 

 new apparatus a base of 20 kilometers can 

 be measured in less time and for less cost 

 than one of a single kilometer with the 

 older forms of measurement. 



THE GREAT AFRICAN ARC OF MERIDIAN 



In connection with the progress of 

 geodesy, time only permits me to say a few 

 words about the Great African are on the 

 thirtieth meridian, which it is a dream of 

 my life to see completed. 



The gap in the arc between the Limpopo 

 and the previously executed triangulation 

 in Rhodesia, which I reported to the Asso- 

 ciation at the Johannesburg meeting in 

 1905, has now been filled up. My own 

 efforts, at 6,000 miles distance, had failed 

 to obtain the necessary funds, but at Sir 

 George Darwin's instance contributions 

 were obtained from this association, from 

 the Royal Society and others, to the extent 

 of half the estimated cost; the remaining 

 half was met by the British South Africa 

 Company. But for Darwin 's happy inter- 

 vention, which enabled me to secure the 

 services of Captain Gordon and his party 

 before the Transvaal Survey Organization 

 was entirely broken up, this serious gap in 

 the great work would probably have long 

 remained; for it is one thing to add to an 

 existing undertaking of the kind, it is quite 

 another to create a new organization for a 

 limited piece of work. 



Since then Colonel (now Sir William) 



