Jan. 8, 1 8 74] 



NA TURE 



185 



of the temperature curve (1779 maximum and 1785 

 minimum) are precisely where, according to the mean 

 length of the sun-spot period of in years, they mu»t be ; 

 that there may, perhaps, be an 11 years period in the 

 temperature independent of the sun-SDOt period, and that, 

 in the prf sent case, a displacement which the spot period 

 has experienced is not shared by the temperature period ; 

 we have to remember that the correspondence of the 

 temperature changes in 1S15 — 54, does not merely ex'end 

 to the average length of the periods, but that all peculiari- 

 ties and disturbances in the sun-spot curve are, in these 

 30 or 40 years, reflected in the temperature curve. Further 

 observation is needed to explain this phenomenon. Pos- 

 sibly (the author suggests), we have here the interference 

 of a number of quite independent periodical actions ; and 

 (without laying stress on the fact, in default of causal 

 evidence), he notices that the greatest negative anom.alies 

 occur, for a considerable time, in a series which progresses 

 by multiples of 9, and in such a manner that an interval 

 of 27 alternates with one of 18 years. Thus — 



1740 = 1767 = 1785 = 1812 = 1S30 = 1857 

 -1-27 +iS -f27 +18 +27 

 The first four agree ; there is merely the quite isolated 

 cold year 1794 intermediate. Going further, we find 

 divergence ; for the table shows a strong negative 

 anomaly about 1836 ; but we have, again, the well- 

 authenticated negative anomaly of 1S56 — 5 7 conforming 

 to the rule. Renou has assigned, for the return of the 

 cold winter of south-western Europe, a period of 41 

 years ; the author asks whether the time 27 :4- 18 = 45 

 years does not better agree with the phenomenon. On 

 this view, the first winter, reckoning back from 1740 is 

 1695, and this is recorded as having been one of excessive 

 cold. Between these two occurs one winter of extra- 

 ordinary cold, 1709, but it is quite isolated, the neigh- 

 bouring years having been warm. If we go still further 

 back, the periodicity cannot be ascertained with any 

 certainty. If the rule is correct, and its validity between 

 1740 and 1857 not a mere accident, i.e. the expression of 

 quire other laws, we have to look for a very cold year in 

 1875 (being 1S57 -f 18). 



Dr. Kiippen proposes, in a future communication, to 

 treat of hydro-meteors, and to examine the influence of 

 periodic weather changes (at several years' interval) on 

 some phenomena of organic nature. 



LAVOISIER'S WORK IN THE FOUNDATION 

 OF THE METRIC SYSTEM 



SINCE the publication of the article on the Metric 

 System, in Nature, vol. viii. p. 3S6, my attention 

 has been drawn to some recent information showing the 

 important part taken by the celebrated Lavoisier in the 

 scientific operations for establishing the basis of the 

 metric system of weights and measures in France. 

 Lavoisier's name has hitherto been little noticed amongst 

 those of the men of science who were prominently en- 

 gaged in this work ; but it is now clearly proved that up 

 to the period of his being guillotined on May 8, 1794, 

 when he fell a victim to the revolutionary fury during the 

 reign of terror, no one took a more active or serviceable 

 part in the scientific labours for founding the Metric 

 System than Lavoisier. 



This information is contained in a " Notice historique 

 sur le Systeme Metrique," by General Morin, lately 

 published in the " Annales du Conservatoire des Arts et 

 Metiers." It is derived from original documents left by 

 Lavoisier, and now in the possession of the Academic des 

 Sciences. These documents have since been submitted 

 to my inspection by M. Dumas, and full details of them 

 will soon be given to the world in the fifth volume of the 

 works of Lavoisier, which M. Dumas is now completing. 



Although Lavoisier's name does not appear in the list 

 of the original Committee of Weights and Measures in 

 France, yet it is shown that he was very actively engaged 

 in making the arrangemen's for th^ir meetings and in 

 preparing ihe minutes of their proceedings, as appears 

 from papers and le ters in his own handwriting. It w s 

 through his personal agency that funds were provided at 

 Paris lor continuing the measurement of the arc of the 

 meridian in Spain by Mdcham. And more particularly, 

 all the actual comparisons for determining the length and 

 dilatation of the standard measures used by Mdchain and 

 Drlambre for measuring the basis, and known as the 

 higles de Borda, were made, not by Borda, but by 

 Lavoisier. The subsequent computations only were 

 made by Borda. Lalande has expressly stated that the 

 work of preparing them was executed by Lavoisier and 

 Borda, but that the construction of the measures of plati- 

 num aijd brass, forming metallic thermometers, and of 

 the comparing apparatus used, was carried out under 

 Lavoisier's directions. The published report upon the 

 construction and verification of these measures in 1792 

 is contained in the " Base du Systeme Metrique," vol. iii. 

 p. 313. It was drawn up by Borda, but Lavoisier's name 

 is not mentioned in it. 



Another very important part of the work, the deter- 

 mination of the weight of a cubic decimetre of water, was 

 carried out, in the first instance, chiefly by Lavoisier. 

 This branch of the operation had been specially entrusted 

 by the Committee to Lavoisier and Haiiy. The necessary 

 apparatus was constructed under Lavoisier's directions, 

 and all the requisite measurements and weighings ot the 

 cyhnder were made by Lavoisier and Haiiy. Hitherto 

 few details of the actual processes of this scientific deter- 

 mination have been given to the public, and the whole 

 credit of determining the weight of a cubic decimetre of 

 water, upon which the kilogram, the unit of metric 

 weight, was based, has been attributed to Lefevre-Gineau, 

 to whom, in conjunction with Fabbroni, the work was 

 entrusted after Lavoisier's death. In point of fact, Le- 

 fevre-Gineau appears to have repeated, in the winter of 

 1798-9, all the observations made by Lavoisier and Haiiy 

 five years before, using the same instruments and obtain- 

 ing nearly similar results. 



The facts are stated as follows by Bugge, the Danish 

 member of the Commission, in the thirtieth of his lexers 

 describing his visit to Paris, and published in 1800 : — 



" The final results of the labours of this special com- 

 mission, consisting of Lefevre-Gineau and Fabbroni, to 

 whom Van Swinden and Trallis were afterwards joined), 

 was that the true kilogram, the weight of a cubic deci- 

 metre of water at its maximum density, or at 4° C, was 

 18,827 French grains of the old French pound, /17/rf'j <fe 

 marc. 



"By the laws of August i, 1763, and April 7, 1795, the 

 kilogram is determined to be 18841 grains of the old 

 French pound, poids de marc, in accordance with the 

 experiments of Lavoisier and Haiiy. This determination 

 was adopted by the Chief Office of Weights and Mea- 

 sures in France, and the Standards have been hitherto 

 made for the Departments accordingly. So that there 

 now exist two kinds of kilograms, the legal or provisional, 

 and the scientific or true kilogram. The difference be- 

 tween them is fourteen old French grains." 



The difference is partly attributable to Lavoisier's de- 

 termination having been made at the temperature of 

 melting ice, instead of that of the maximum density of 

 water adopted tor Lefevre-Gineau's determination. The 

 unit of Metric weight, the Kilogramme des Archives, ap- 

 pears to have been based on the later observations of 

 Lefevre-Gineau, and to have been legalised by the law of 

 Dec. 9, 1799, after Bugge's letter was written. 



H. W. Chisholm 



