i86 



NATURE 



{Dec. 26, 1889 



in other direc'ions, and could then concentrate its efforts on what 

 is perhaps, after all, its most legitimate and most useful function 

 — that of providing intermediate technical schools on the pattern 

 of the Finsbury School, of which many are required in the 

 metropolis. 



The exact terms on which the Government would be prepared 

 to take over this part of your work is a subject on which, of 

 course, I cannot pretend to enter, but a satisfactory basis can, I 

 do not doubt, easily be found. 



Your Council would then feel that the great work which they 

 have begun has been handed over in its full vigour to the nation, 

 and that with the nation lies the responsibility of extending and 

 perfecting the system which they have had the honour and the 

 gratification of inaugurating, 



I am aware that in making these suggestions, I have raised a 

 somewhat burning question about which there may be difference 

 of opinion, and my apology for this indiscretion, if one is 

 needed, must be the importance of the subject, and the anxiety 

 which we all feel that the technical education of our country 

 shall be placed on a firm and enduring national basis. 



A FIRST FORESHADOWING OF THE 

 PERIODIC LA W. 



TT is well known that the Newlands-Mendeleeff classification 

 of the elements was preceded by the discoveries of certain 

 numerical relations between the atomic weights of allied ele- 

 ments, due to Dobereiner, Dumas, and others ; but what has 

 been almost entirely ignored is the immense advance made by 

 M. A. E. Beguyer, de Chancourtois,^ a French geologist of note. 

 Professor at the Ecole des Mines, who was the first to publish 

 a list of all the known elements in the order of their atomic 

 weights. 



M. de Chancourtois embodied his results in two memoirs 

 presented to the French Academy of Sciences in April 1862 

 and March 1863. These memoirs have never been printed in 

 ^xtensoj^ but extracts from them, and additional notes relating 

 to the subject, were published in the Comptes rendus for 1862 

 (liv. pp. 757, 840, and 967 ; Iv. p. 600), 1863 (Ivi. pp. 253 

 and 479), and 1866 (vol. Ixiii. p. 24). The first note bears 

 the date of April 7, 1862, so that there can be no doubt as to 

 de Chancourtois's claim to priority in this important matter.^ 



I have in my possession a thin quarto pamphlet, by de 

 Chancourtois, entitled " Vis Tellurique : Classement naturel des 

 corps simples ou radicaux, obtenu au moyen d'un systeme de 

 classification helicoidal et numerique " (Paris, Mallet-Bachelier,* 

 1863), which contains nearly all the extracts from the Comptes 

 rendus, together with some additional matter. It contains, also, 

 what is absolutely essential to the comprehension of de Chan- 

 courtois's ideas, the graphic representation of his system, which 

 is not to be found in the Comptes rendus. 



I propose to give here a translation of the first communica- 

 tion to the Academy, followed by certain explanatory comments 

 and brief extracts from the other papers : — 



"Geological studies in the field of research opened up by 

 M. Elie de Beaumont in his note on volcanic and metalli- 

 ferous intrusions {emanations) have led me, for the completion 

 of a lithological memoir on v^'hich I am now engaged, to a 

 natural classification of the simple bodies and radicles by a table 

 in the form of a helix, founded on the use of numbers which I 

 ■call characteristic numbers or numerical cliaracteristics. 



" My numbers, which are immediately deduced from the 

 measure of the equivalents or other physical or chemical capacities 

 of the different bodies, are, in the main, the proportional numbers 

 ^iven by the treatises on chemistry, these being reduced to half 

 in the case of hydrogen, niirogen, fluorine, chlorine, bromine, 



I Wurtz("The Atomic Theory," p. 170) and Berthelot (" Les Origines 

 •de rAlchimie," p. 302) give a bare mention of de Chancourtois's name. 

 Mendeleeflf, in his Faraday Lecture (Journ. Chem. Soc, October 1889), 

 couples his name with those of Newlands and Strecker, and shows greater 

 appreciation of his work. 



^ M. Friedel, the eminent ProfessDr of Organic Chemistry at the Sorbonne, 

 has kindly procured for me the information that the original manuscripts of 

 these memoirs are preserved in the archives of the Institut ; I hope to be 

 able to examine them at some future period. 



3 Hr. Newlands' first paper, chiefly devoted to showing that the nume- 

 rical differences between the atomic weights of allied elements are approxi- 

 mately multiples of 8 was publ.shed on February 7, 1863 {Chemical News, 

 vol. vii. p. 70) ; his second paper, in which he arranges the elements in the 

 order of their atomic weights, was published on July 30, 1864 {Chemical 

 News, vol. X. p. 39) Sej J. A. R. Newlands "On the Discovery of the 

 Periodic Law," Ike. (Spon, 1884). 



4 Now Gauthier-Villars. 



Iodine, phosphorus, arsenic, lithium, potassium, sodium, and 

 silver ; in other words, I either divide the equivalents of these 

 bodies by two in the system in which oxygen is taken as 100, 

 or multiply by two the equivalents of the other bodies in the 

 system in which hydrogen is taken as unity. 



" On a cylinder with a circular base, I trace a helix which cuts 

 the generating lines at an angle of 45°. I take the length of one 

 turn of the helix as my unit of length, and starting from a fixed 

 origin, I mark off on the helix lengths corresponding to the 

 different characteristic numbers of the system in which the 

 number for oxygen is taken as unity. The extremities of the 

 lines thits marked off determine points on the cylinder which I 

 call indifferently characteristic points or geometrical cliaracters, 

 and which I distinguish by the ordinary symbols for the different 

 bodies. The same points will evidently be obtained if we take 

 as the unit of length the ^V of ^ turn of the helix, and mark off 

 on the curve lengths corresponding to the numbers of the system 

 in which hydrogen is represented by unity. 



"The series of points thus determined constitutes the graphic 

 representation of my classification, which may easily be traced 

 on a plane surface by supposing the stirface of the cylinder de- 

 veloped ; by its aid I am enabled to enounce the fundamental 

 theorem of my system: The relations between the properties of 

 different bodies are manifested by simple geotnetrical relations 

 between tlie positions of their characteristic points. 



" For instance, oxygen, sulphur, selenium, tellurium, bismuth,^ 

 fall approximately on the same generating line, while magnesium, 

 calcium, iron, strontium, uranium, and barium, fall on the 

 opposite generating line. On either side of the first of these 

 lines we find hydrogen and zinc on the one hand, bromine and 

 iodine, copper and lead on the other ; parallel to the second line 

 we find lithium, sodium, potassium, manganese, &c. 



" Simple relations of position on a cylindrical surface would be 

 obviously defined by means of helices, of which the generating 

 lines are only a particular case ; hence, as a complement to the 

 first theorem, we may add the following : Each helix drazvn 

 through two characteristic points and passing through several 

 other points or only near them, brings out relations of a certain 

 kind between their p7-operties ; likenesses and differences being 

 manifested by a certain numerical order in their succession, for 

 example, immediate sequence or alte?-nation at various periods. 



" In order to attain a greater degree of accuracy, it is necessary 

 to discuss the results of different measurements with respect to 

 the same body. 



"One question is all-important in this discussion; it is to 

 know if the divergencies which occur may have causes other 

 than the error of experiment. I reply to this question in the 

 affirmative. 



" I think that here, as in all determinations of constants which 

 we wish to compare, they must be reduced to the same con- 

 ditions. This idea seems to me the indispensable complement 

 to the notion of an absolute characteristic number. Once the 

 existence of this absolute number or numerical characteristic 

 guaranteed by the possibility of connecting it afresh with ob- 

 served facts, certain limits of variation being allowed {literally, 

 varying within certain limits], we promptly arrive at Front's 

 law, which presents itself as furnishing a means for reducing 

 experimental observations to a comparable state by a series of 

 trials, without this state being even a completely defined one, 

 but, on the contrary, in order to be able to define it. The 

 combination of this principle with the rules for alignment allow 

 me to give the most striking form to my invention. I am thus 

 led to formulate the table of integral numbers, which, as I must 

 not omit to mention, exhibits under certain aspects the rSsumS 

 of the work of M. Dumas on this subject. 



" In the construction of this table I have had recourse to the 

 determinations of specific heats, not only as a means of control, 

 but also to find new numbers unattainable by the methods of 

 chemical investigation. By adopting as the constant product of 

 specific heat by atomic weight, the number which corresponds 

 both to sulphur and to lead, I have deduced from the series of 

 results given by M. Regnault, purely thermic quotients or num- 

 bers, which take their places on my alignments in the most felici- 

 tous way. I will only quote two examples : firstly, the number 

 44, obtained from the specific heat of the diamond, which finds 

 its place on the generating line of the characteristic, 12, of car- 

 bon, by the side of the characteristic, 43, which corresponds to 

 one of the equivalents generally accepted for silicon ; and another 



' This is probably a misprint, as bismuth does jiot fall on the same 

 generating line in the table. 



