i88 



NA TURE 



[Dec. 26, 1889 



Three important points, however, do exist in common 

 between de Chancourtois's system and that of Mendeleeff: — 



Firstly, all the known elements are arranged in the order of 

 their combining weights. 



Secondly, the combining weights chosen as best suited to 

 bring out clearly the numerical relations existing between them 

 are those adopted by Cannizzaro in 1858, a striking fact when 

 we recollect that de Chan'courtois wrote only in 1862, at a date 

 long before these numbers had gained anything like general 

 acceptance. 



Lastly, an attempt is made to show that simple numerical 

 relations exist, not only between the combining weights, but 

 between all the measurable properties {toiites les capacitcs 

 physiques et chimiques') of allied elements. 



The reasons for the neglect of de Chancourtois's researches 

 and the oblivion into which they have fallen are not far to seek. 

 His style was heavy and at times obscure, and, moreover, his 

 ideas were presented in a way most unattractive to chemists. 



A geologist by profession, de Chancourtois had been power- 

 fully impressed by the facts of isomorphism in the case of the 

 feldspars and pyroxenes, which form such important constituents 

 of the volcanic rocks he was studying ; and he was thus led to 

 seek for a system of classification which should bring out some 

 simple relationship between the elements they contained. 



To quote from his paper {Comptes rendus, vol. liv. p. 969) : 

 "The parallelism of the groups of manganese (7 + 3 . 16) 

 and iron (8 + 3 . 16), of potassium (7 + 2 . 16) and calcium 

 (8 + 2 . 16), of sodium (7 + 16) and magnesium (8 + 16), is the 

 origin of my system " ; and again, suggesting the expediency of 

 adopting 55 (= 7 + 3 . 16) as a characteristic for aluminium, 

 which would bring the element on the sodium and potassium 

 generating line, "this would render perfect the parallelism 

 between the elements of the feldspars and the pyroxenes, the 

 starting-point of my system " {Comptes rendus, Ivi. p. 1479)- 



Thus the correct idea of seeking for a relationship between 

 the combining weights of isomorphous elements was marred 

 by a somewhat imperfect comprehension of the facts of 

 isomorphism. No chemist would certainly have tried to show 

 any close relationship between aluminium on the one hand and 

 the group of the alkalies on the other, notwithstanding their 

 union in the feldspars and pyroxenes ; and a suggestion of this 

 kind served to cast discredit on de Chancourtois's really important 

 views. 



Notwithstanding his frequently eccentric ideas, de Chancour- 

 tois had the merit, so rare in an inventor of this stamp, of not 

 considering his system as final. We cannot do better than let 

 him speak for himself ; and quote the conclusion of his last paper 

 on the subject (Comptes rendus, Ivi. p. 481) : — " In presence 

 of the rapid increase in the list of elements which engage the 

 attention of chemists and physicists, it has become urgent to 

 unite in one synthesis all the notions of chemical and physical 

 capacities, of which the exposition would otherwise become an 

 impossible task. 



" It is, therefore, perhaps not unnecessary to recall the ideas of 

 Pythagoras, or what I may better term the Biblical truth which 

 dominates all the sciences, and of which I propose to make 

 practical use by the following concrete example,^ the first general 

 conclusion of my essay : — 



"The properties of bodies are the properties of 



NUMBERS. 



"It is easily perceived, that a helicoidal system of some kind 

 or another, which is necessarily a graphic table of divisibility, 

 •offers the most convenient means for rendering manifest the 

 relations between the two orders of ideas. It is evident, also, 

 that the particular system which I have adopted brings into 

 relief the relations of the most important and usual of the proper- 

 ties of matter, because the case of divisibility by 4, which is the 

 basis of my plan, is the first which presents itself in arithmetical 

 speculation after the case of divisibility by 2, to which there 

 •corresponds directly, as one perceives by a first glance at my 

 table, the existence of the natural couples of elements, with 

 consecutive odd and even characteristics. 



" I hope, therefore, that the telluric helix will offer, until it 

 is replaced by some more perfect invention, a practical frame- 

 work, a convenient scale, on which to set down and compare all 

 measurements of capacities, whatever the point of view which 

 may be taken, whatever elasticity or variation, whatever inter- 

 pretation may be given to the nuinerical characteristics, by which 

 these capacities must always be represented. 



■* The French is vulgarisation, WicraWy j>oJ>Hlarixatitn. 



" The development in a plane of the cylinder ruled 

 into squares, with the circumference at the base divided into 

 16 equal parts, seems to me, in a word, to be a stave on 

 which men of science, after the fashion of musicians, will note 

 down the results of their experimental or speculative studies, 

 either to co-ordinate their work, or to give a summary of it in 

 the most concise and clear form to their colleagues and the 

 public. " 



Lothar Meyer has noted down his classification in the form of 

 a helix, '^ and Dr. Johnstone Stoney, F.R.S., has shown that the 

 numerical values of the atomic weights may be expressed geo- 

 metrically as functions of a series of integral numbers by points 

 all lying approximately on a logarithmic spiral. 



But no simple mathematical formula has so far been discovered 

 to express the relationships of the atomic weights accurately — 

 i.e. within the limits of experimental error, and de Chancourtois's 

 predictions still remain but incompletely fulfilled. 



I need not comment further on the remarkable breadth and 

 originality of our author's views, taken as a whole. Strange to 

 say, it was only a year or two before his death that he heard, 

 through a colleague, of the immense development they had 

 undergone ; nor did he ever set up any claims to priority. But 

 when we speak of the greatest generalization of modern chemistry, 

 and recall the names of Newlands and Mendeleeff, it is only just 

 that we should no longer forget their distinguished precursor, 

 de Chancourtois. P. J. Hartog. 



SCIENTIFIC SERIALS. 



American [ournal of Science, December. — The temperature 

 of the moon, by S. P. Langley, with the assistance of 

 F. W. Bery. With this memoir the authors complete the 

 researches begun at the Allegheny Observatory in 1883 and 

 continued during the next four years. The main outcome is 

 that the mean temperature of the sunlit lunar surface is much 

 lower than has been supposed, most probably not being greatly 

 above 0° C. — The Lower Cretaceous of the South- West, and its 

 relation to the underlying and overlying formations, by Charles 

 A. White. The chalk formations constituting the so-called 

 " Texas Section " are here referred to two natural divisions, 

 which may be designated the Upper and Lower Cretaceous 

 respectively, although not necessarily the exact equivalents of 

 the corresponding European strata. Their fossil contents show 

 that each represents an unbroken portion of Cretaceous time, 

 while the palaeontological contrast between the two indicates 

 that there is a time hiatus between them. But this hiatus is no 

 greater than exhibited in others of the mountain uplifts in the 

 same region, and not so great as it is in some cases. — On the 

 hinge of Pelecypods and its development, with an attempt 

 toward a better subdivision of the group, by William H. Dall. 

 Three fundamental types of hinges are described, and on these 

 is based a new classification comprising the three orders of 

 Anomalodesmacea with five sub-orders, Prionodesmacea with 

 eight sub-orders, and Teleodesmacea with eleven or more sub- 

 orders. — The magnetism of nickel and tungsten alloys, by 

 John Trowbridge and Samuel Sheldon. The question is here 

 discussed whether nickel and timgsten alloys magnetized to 

 saturation increase in specific magnetism as different kinds of 

 steel alloyed in small proportions with tungsten or wolfram are 

 known to do. The tabulated results show that tungsten greatly 

 increases the magnetic moment of nickel, if the alloy be forged 

 and rolled, but has small influence if simply cast ; nor do changes 

 in the amount of tungsten appear to cause corresponding changes 

 in the magnetic properties of the alloy. — Note on the measure- 

 ment of the internal resistance of batteries, by B. O. Peirce 

 and R. W. Willson. The authors' researches show that the 

 value of the resistance of a cell obtained by the use of alternate 

 currents is always smaller than that obtained by other methods, 

 but the application of the method of alternate currents " fatigues " 

 all but the so-called constant cells. In most cases there is a 

 tendency in the internal resistance to decrease as the strength of 

 the current which the cell is delivering increases. — Papers were 

 contributed by Robert T. Hill and R. A. F. Penrose, Jun., on 

 the relation of the uppermost Cretaceous beds of the Eastern 

 and Southern United States, and on the Tertiary Cretaceous 

 parting of Arkansas and Texas ; by W. E. Hidden and 



^ "Die modernen Theorien der Chemie," iv. Auflage, p. 137; English 

 translation, p. 118. 



