UNIVERSITY OF CALIFORNIA LIBRARY LIBRARY OF THK UNIVERSITY OF CALIFORNIA. Class HARPER'S LIBRARY of LIVING THOUGHT THE ELEMENTS SPECULATIONS AS TO THEIR NATURE AND ORIGIN BY SIR WILLIAM A. TILDEN HARPER BROTHERS LONDONXNEWYOKK THE ELEMENTS SPECULATIONS AS TO THEIR NATURE AND ORIGIN BY SIR WILLIAM A. TILDEN F.R.S., D.SC. LOND., HON. SC.D. DUB., D.SC. VIC., LL.D. BIRM. Fellow of the University of London. Past President of the Institute of Chemistry and of the Chemical Society, Late Professor of Chemistry in the Royal College of Science and Royal School of Mines. Davy Medallist of the Royal Society, Honorary Member of the Pharmaceutical Society of Great Britain, of the Society of Public Analysts, of the Bristol Society oj Naturalists, of the Birmingham Philosophical Society and of the Philadelphia College of Pharmacy, etc. Professor Emeritus in the Imperial College of Science and Technology. LONDON AND NEW YORK HARPER & BROTHERS 45 ALBEMARLE STREET, W. 1910 PREFACE \ 7[ 7HEN in October last, by invitation * of the Council of the Chemical Society, I gave the Mendeleeff Memorial Lecture, it was my privilege to proclaim, on behalf of English science, a eulogium of the famous Russian chemist. It would have been inappropriate to the occasion to have discussed in much detail the various hypo- theses which from time to time have been framed in connection with the idea of evolution as applied to the elements, and all the more so for the reason that this idea seems to have been always repugnant to the mind of Mendeleeff himself. I am therefore glad of the opportunity afforded by this series to place side by side ideas which have been long fermenting in my own mind and in the minds of other chemists. In doing this I do not pretend to have treated vii 216668 PREFACE the subject with any approach to historical completeness. The literature connected with the subject is very extensive, and is still expanding. What I have tried to do is to render the discussion intelligible to the very large number of educated persons possessing an elementary knowledge of chemistry, as well as to offer a few suggestions to expert chemists. In doing this I have had to exercise my own judgment in the selection of those contributions to the inquiry which seemed to possess the greatest importance. I am conscious that this lays me open to criticism, alike from those whose views have been discussed and from those who have not been mentioned, as well as from those who see points of attack in the exposition of my own notions. Such, however, is always the position of one who ventures to enter a field so open to specula- tion as the subject of this little book. Here we are dealing only with the physical PREFACE view of the phenomena observed or discov- ered. Concerning the metaphysical view, we must accept the dictum of Herbert Spencer : " Matter, in its ultimate nature, is as abso- lutely incomprehensible as Space and Time. Whatever supposition we frame leaves us nothing but a choice between opposite absurdities." (First Principles, vi. ed. (1900), p. 46.) I cannot conclude without expressing my best thanks to Sir William Crookes for his kindness in allowing the use of the block for the diagram on page 74. W. A. T. April, 1910. IX CONTENTS CHAPTER PAGE I. THE ELEMENTS . l II. INTERRELATIONS AMONG ATOMIC WEIGHTS . 2 5 III. THE PERIODIC LAW . 5 1 IV. THEORIES OF EVOLUTION . 71 V. SPECULATIONS . . . 108 INDEX . . ! 37 XI THE ELEMENTS CHAPTER I " CHEMICAL analysis and synthesis go no farther than to the separation of particles one from another and to their reunion. No new creation or destruction of matter is within the reach of chemical agency. We might as well attempt to introduce a new planet into the solar system, or to annihilate one already in existence, as to create or destroy a particle of hydrogen. All the changes we can produce consist in separating particles that are in a state of cohesion or combination, and joining those that were pre- viously at a distance." DALTON'S Chemical Philosophy (1808), vol. i, p. 212. THE ancient Greek philosophers, and probably others in earlier times, from the contemplation of the order of things unfolded before them, were led to the dis- cussion of problems which could never have been resolved by the methods alone avail- able in the early ages of the world. The substitution of artificial for natural conditions in the study of phenomena, in B I THE ELEMENTS other words the use of experiment, is the characteristic of modern philosophy, and without it the modern student would be in nearly the same position as Democritus and Leucippus, and would doubtless remain in the same dialectical obscurity. But experimental physics and chemistry, comparatively recent as they are in their origin, have led us far beyond anything which in ancient times could have been con- ceived as possible in respect to positive knowledge of the constitution and order of the visible universe. Questions as to the nature and origin of matter can therefore be treated at the present day as within the range of reasonable subjects of study, to which answers can be framed in terms which, if not final, are at any rate general, consistent, and intelligible. Chemistry teaches us that all the immense diversity of matters which go to make up the solid earth, with its attendant ocean and atmosphere, together with the sun and all DEFINITION OF ELEMENT the heavenly host of stars and nebulae, are composed of a limited number of distinct substances which are called " elements."* * Perhaps it may not be superfluous to remind the reader that the word "element" has received at different periods in the history of philosophy several distinct applications. The four elements fire, water, earth, air of the Aristotelean system, were succeeded in the sixteenth century by the salt, sulphur, and mercury, or the tria prima of the alchemists. To Robert Boyle we owe the connotation now universally attached in scientific language to this word, which is applied to those substances, about eighty in number, from which, by the operation of ordinary chemical processes, only one kind of matter can be obtained. Iron, for example, is placed among the elements, while iron rust is a compound of two elements, iron and oxygen. Notwithstanding the belief now generally entertained that some of the so-called elements subsist in a condition of continuous decay, this application of the word is convenient and is likely to be retained for some time to come. The state of knowledge and opinion in Boyle's time may be inferred from the following caustic passage in his famous book entitled, The Sceptical Chymist" (1680), pp. 23, 24. " The doctrine of the four Elements was framed by Aris- totle after he had leasurely considered those Theories of former Philosophers. . . . Nor has an Hypothesis so deliberately and maturely established been called in Ques- tion till in the last Century Paracelsus and some few other sooty Empiricks rather than (as they are fain to call them- selves) Philosophers having their eyes darken'd and their Braines troubl'd with the smoak of their own Furnaces, began to rail at the Peripatetick Doctrine, which they were too illiterate to understand, and to tell the credulous World THE ELEMENTS Many of these are familiar in daily life. Such metals as pure gold, silver, iron, copper, lead, tin, zinc, etc., are elements, as also are the atmospheric gases oxygen, nitrogen, argon. These are for the most part capable of associating together in chemical " com- pounds," in which the properties of the " elements," as commonly known, are con- cealed and substances of different aspect and properties result. Water, for example, is a very different thing from the oxygen and hydrogen into which by appropriate treat- ment it can be wholly resolved. The change of a compound into elements is always accom- panied by a redistribution of the energy inherent in the matter concerned and in the matter with which it is in contact, or in the that they could see but three Ingredients in mixt Bodies ; which to gain themselves the repute of Inventors they endeavoured to disguise by calling them instead of Earth and Fire and Vapour, Salt, Sulphur, and Mercury ; to which they gave the canting title of Hypostatical Principles, but when they came to describe them they showed how little they understood what they meant by them by disagree- ing as much from one another as from the truth they agreed in opposing." 4 DIVISIBILITY OF MATTER " ether "* in which all are immersed. In the majority of cases the separation of elements from a compound is attended by absorption of energy, which is given out again in the form of heat if they are allowed or caused to combine. Water, for example, when strongly heated gives oxygen gas and hydrogen gas, and these if allowed to combine again re- produce water, at the same time giving forth heat equivalent in amount to the energy used up in the destruction of the compound. The old question whether the divisibility of matter is finite or infinite, debated throughout ancient and medieval times without the possibility of reaching a con- clusion, has been in modern times decisively answered by physics and chemistry. It is no longer a subject of debate, but it is with practical unanimity agreed that the texture of all kinds of matter is not con- tinuous, but is discrete or granular. Hence * See Lodge, The Ether of Space. This series. 5 THE ELEMENTS the Atomic and Molecular Theory, which asserts that there is a limit to divisibility, and the coarseness or fineness of the atomic particles is merely a detail. The nature of the evidence on which this conclusion is based is derived wholly from modern experimental investigations, on the one hand, of the properties of liquids and especially of gases ; and on the other, from the establishment of the fundamental quanti- tative laws of chemical combination. The facts of gaseous and liquid diffusion are familiar, and alone they are sufficient to prove that in a mass of any fluid portions of it are constantly moving about, and that these moving portions are very small is indicated by their passage through the pores of earthenware, or even of various mem- branes. A hundred years ago it was discovered that when two substances unite to form a chemical compound the ratio between the quantities of the two so uniting is constant. 6 DEFINITE PROPORTIONS That is to say, any selected chemical com- pound consists of two or more elements associated together in proportions which never vary. Water, for example, is always composed of hydrogen and oxygen in the proportions of i part by weight of the former to 8 parts by weight of the latter, and no kind or variety of water is known in which the two components are united in different proportions. There is indeed an- other compound of hydrogen with oxygen, in which the weight of the oxygen is sixteen times that of the hydrogen, but this com- pound is wholly different from water in all its properties. It will be noticed that the proportion of oxygen in the second com- pound is exactly twice the proportion in the former. There are many examples of the same kind, and in all such cases when two substances unite in several proportions it will be found that the proportion of one element in the successive compounds are multiples of the proportion of this element 7 THE ELEMENTS occurring in the first of the series. Com- pounds are known containing 28 parts of nitrogen combined with 16 parts of oxygen, and with twice, three times, four times, and five times 16 parts of oxygen in the successive stages. This is a simple illustration of the opera- tion of the law of multiple proportions, and for this no satisfactory explanation has been given except that which is furnished by the Atomic Theory of Dalton. Dalton's theory was announced in his New System of Chemical Philosophy, which he published in 1808. But for many years the system made little progress owing to the fact that on the one hand Dalton and his contemporaries had no standard by which the relative weights of the atoms of different elements could be determined ; and on the other hand because a distinction in terms, generally acceptable to physicists and chemists, had not been established between the ultimate particles of ele- 8 ATOM AND MOLECULE ments and of compounds. While it is perfectly justifiable to speak of an atom of water, meaning thereby the particle which if further subdivided would be no longer water, but would become a mixture of hydrogen and oxygen, yet this use of the word atom is liable to lead to confusion, and another term was required. This was pro- vided later by the use of the word molecule, which signifies the smallest particle of an element or of a compound capable of in- dependent existence, the term atom being reserved for the smallest portion of an element ever found in a molecule. The conception of the molecule as distinct from the atom we owe to Avogadro, who, in 1811, put forward a hypothesis which has only in much later times been generally accepted by chemists. A famous French chemist, Gay-Lussac, established by experi- ment the law with which is generally associ- ated his name, namely, that when gases unite together the volumes which unite stand to 9 THE ELEMENTS one another in a simple ratio. For example, one volume of hydrogen combines with one volume of chlorine, producing two volumes of hydrogen chloride gas, one volume of hydrogen chloride combines with one volume of ammonia, two volumes of hydrogen com- bine with one volume of oxygen, etc. In order to explain this discovery Avogadro assumed that equal volumes of different gases, at the same temperature and pressure, contain the same number of molecules, a statement which is often expressed other- wise by saying that the densities of gases are proportional to the molecular weights of the substances of which they consist. If this be accepted then a means is pro- vided whereby atomic weights may be deter- mined, and for all those elements which are capable of yielding gaseous or vaporisable compounds the atomic weights so deter- mined are comparable with one another and with a common standard. All that is fur- 10 STANDARD FOR ATOMIC WEIGHTS ther necessary is to agree upon some one element the atomic weight of which may be adopted as the unit. For this purpose hydro- gen, as forming the lightest gas known, and as the element which enters into combination in the smallest known proportion, has during the last half century been accepted. Latterly for reasons which at this point need not be considered, the scale of hydrogen as the unit, =i, has been, by international con- sent, exchanged for one in which the atomic weight of oxygen is taken as exactly 16, and consequently that of hydrogen becomes i -008. The Atomic Theory is further supported by the remarkable achievements in the do- main of stereochemistry which have been recorded during the last thirty years. Not- withstanding the large number of hypo- theses which have been put forward no explanation is yet established of that property of atoms which is called their valency. But though it is not possible to ii THE ELEMENTS explain what is the nature of the link by which atoms unite together in chemical com- bination, it is possible by the aid of the assumption that matter consists of atoms united into groups or clusters, called mole- cules, and the further assumption that when these atoms occupy in space certain positions relatively to one another, to show that the molecules so constituted exhibit certain recognisable properties. It is now a familiar operation to proceed to build up such con- geries of atoms with the certainty that the resulting compound will exhibit the ex- pected properties. And, further, it has been shown in a large number of cases that, given a limited number of atoms, a limited number of arrangements are alone possible, and any attempts to pro- duce other compounds in which the rules imposed by valency are not complied with result in failure. This is not the place for an exposition of the results of modern stereochemical in- 12 STEREOISOMERIC COMPOUNDS vestigation, but for the sake of illustration a single familiar example may be quoted. There are four kinds of tartaric acid, namely : (i) common tartaric acid which rotates the plane of polarisation of light toward the right ; (2) another variety which rotates to the left ; (3) racemic acid, which has no rotatory power, but is optically inactive be- cause it consists of a mixture in exactly equal quantities of the two active varieties which are separable by simple processes from each other ; and (4) an optically inactive form of tartaric acid which is not resolvable into two active forms. These acids consist of carbon hydrogen and oxygen in the proportions re- presented by the formula C 4 H 6 O 6 , and they agree in chemical reactions which are ex- pressed by saying that they are dihydroxy- succinic acids, and hence that they consist of the atomic clusters or radicles repre- sented by the symbols C 2 H 2 , 2HO, and 2CO 2 H. Two of these acids act on polarised light 13 THE ELEMENTS in such a way as to show that their molecules must be unsymmetrical in opposite senses, the one corresponding to a right-handed spiral, the other to a left-handed spiral. In racemic acid the dextro-rotatory effect of one of these molecules is exactly neutralised by the lavorotatory effect of the other. The other inactive compound must be accounted for in a different way, namely by the assump- tion that the internal parts of each molecule are so arranged that the lopsidedness pro- duced by one constituent is counterbalanced by another so situated as to act in the opposite direction relatively to the centre of gravity. The phenomena exhibited by the tartaric acids were discovered by Pasteur, about 1850, and he anticipated to some extent the hypotheses introduced and accepted twenty-five years later. But it was not till the idea of attributing to the atom of carbon a peculiarity of configuration was published almost simultaneously by Le Bel and Van't '4 THE CARBON ATOM Hoff * that the atomic and molecular hypo- theses became qualified to explain all the facts. The four units of valency of an atom of carbon being supposed to act only in certain directions represented by the straight lines drawn from the centre of a regular tetrahedron to its solid angles, it was found possible to account for the right- and left- handed optical properties of the tartaric acids and the many other similarly active compounds. It might of course be said that in making use of this conception one hypothesis is employed to support another, but the facts which fit in with the theory are so numerous and its use has led to so many discoveries of compounds, the existence of which was previously unsuspected, that supposition is changed into conviction that this is really * "Sur les formulas de structure dans 1'espace," J. H. Van't Hoff in Archives Neerland, ix (1874), p. 445. "Sur les relations qui existent entre les formules atomiques des corps organiques, et le pouvoir rotatoire de leurs dissolutions," J. A. Le Bel, Bulletin Soc. Chim. y Paris, xxii (1874), P- 337- 15 THE ELEMENTS the physical explanation of the observed facts. Further support for the theory is derived from the results of applying to the com- pounds of elements other than carbon similar methods of investigation. For it has been found that nitrogen, sulphur, tin, and probably some of the metals are also cap- able of giving rise to isomeric compounds of which the optical and other physical relation- ships finds 'a corresponding explanation. ""Having briefly reviewed the chief con- siderations which have led to the adoption of the Atomic Theory, it is proper to survey equally briefly the facts and arguments by the aid of which chemists have finally arrived at a system of numbers which represent the relative masses of the atoms of elements. These numbers are really ratios or fractions in which the denominator is suppressed, being understood to be throughout the value of the atomic weight of the element hydrogen, the smallest at present known. In Dalton's 16 STANDARD FOR ATOMIC WEIGHTS time, and long afterwards, the numbers called atomic weights were in reality the chemical combining ratios of the elements, and these might or might not coincide with the numbers which are now chosen in accordance with the results of the applica- tion of rules agreed upon since that day. It is not necessary to trace all the various suggestions put forward during nearly half a century with the object of reducing these figures to order. It is sufficient to say that to Cannizzaro is due the credit of success- fully convincing the chemical world of the desirability of adopting a uniform standard, and of bringing to a focus the proposals of other chemists, especially those of Gerhardt, Odling, Williamson, and others, based on " the corner-stone of the modern theory of molecules and atoms/' * the theory of Avogadro on the constitution of gases. The process by which an atomic weight is * Cannizzaro : Faraday Lecture, 1872. Trans. Chem. Sof. t 25, 946. C 17 THE ELEMENTS determined resolves itself into two distinct parts, of which the first is dependent upon accuracy in experiment ; the second on the selection of appropriate theoretical con- siderations. The first requisite in all cases is the deter- mination of the chemical combining ratio of the element, or what is frequently, though not very correctly, called the equivalent. This is accomplished by estimating, with every precaution to secure accuracy, either the proportions in which two elements unite together, or the amount set free by the process of electrolysis from a solution of the compound and a comparison of the weight thus deposited with the weight of some standard substance, such as silver, set free at the same time by the same current. The proportions in which oxygen and hydrogen unite to form water have been repeatedly investigated, as also the composition of hydrogen chloride, and the chlorides and oxides of many metals, as well as non-metals. 18 AVOGADRO'S RULE In fact, whenever a compound can be ob- tained in a pure definite state, in which it can be readily weighed, its composition has been the subject of laborious investigation, and thus a series of numbers have been arrived at which represent with a greater or less degree of accuracy combining weights of all the elements. The second step is to choose a multiple of the number representing the combining weight, such that the product complies with one or all of the following rules. i. Application of the rule of Avogadro : The atomic weight of an element is the smallest quantity ever found in two volumes of the vapour of any of its vaporisable compounds, the bulk of one part by weight of hydrogen being taken as one volume. Practically this amounts to comparing the densities of the vapours of the several compounds containing the element in ques- tion with the density of hydrogen as the 19 THE ELEMENTS unit. The results in the case of carbon may be tabulated as follows : Compound Density Density x 2 Weight of Carbo containing Carbon. H i. or weight in, 2 vols of 2 vols. Marsh gas . 8 16 12 Ethane 15 30 24 Carbon monoxide 14 28 12 Carbon dioxide . 22 44 12 Alcohol 23 46 24 Ether . 37 74 48 Aniline 46-5 93 72 The atomic weight of carbon is taken as 12, because this is the smallest weight of carbon ever found in two volumes of the vapour, that is in a molecular weight. 2. Application of the rule of Dulong and Petit. While the relation of specific heat to atomic weight is expressible in the great maj ority of cases by the equation S.H. xA.W. = constant =6 -4 (approx.) this does not hold good save under exceptional conditions for carbon and several other elements. But 20 DULONG AND PETIT it is especially useful in connection with the metals, many of which do not form volatile compounds, and hence cannot be tested by the hypothesis of Avogadro. It is, however, most important to notice that in those cases, tin, mercury, zinc, for example, in which both methods can be applied, the atomic weight deduced from the application of one rule is identical with that which is deduced from the application of the other. 3. The observation of isomorphism is often a useful guide. The case of vanadium furnishes an interesting example. Fifty years ago this element was supposed, on the authority of analyses made by Berzelius, to belong to the same family as chromium, and its highest oxide was represented as a trioxide. But the mineral vanadinite, con- sisting of lead vanadate and chloride, was shown to crystallise in the same form as apatite or calcium fluo-phosphate, and pyromorphite or lead chlorophosphate, and in some cases to crystallise with these 21 THE ELEMENTS minerals in all proportions. Hence it ap- pears that vanadic acid is the crystallo- graphic representative of phosphoric acid, and vanadic oxide like phosphoric oxide is a pentoxide. The whole chemical history of the compounds of vanadium was shown by Roscoe to conform to this view, and hence the atomic weight of vanadium was shown to be approximately 51, instead of 137 as previously supposed, and the substance represented by Berzelius as the metal turned out to be an oxide. 4. The position of the elements in the periodic scheme, to be described later, has been frequently turned to account within the last forty years. Beryllium (glucinum) affords an illustration of the application of this principle. Formerly supposed to be related to aluminium this metal was repre- sented as forming a sesquioxide, Be 2 O 3 , and having the atomic weight 13 (approx.). By reference to the table (p. 46) it ap- pears that between carbon (at. wt. 12) and 22 USE OF PERIODIC SCHEME nitrogen (at. wt. 14) there is no appropriate place for an element with the properties of a metal, nor indeed for any element whatever, if the scheme truly represents the mutual relationships of the ele- ments. This remark led to further investiga- tion of the physical and chemical properties of beryllium with the result that, from ob- servation of the specific heat of the metal at various temperatures and a study of the characters of its salts, this element was shown to be related not to aluminium, but to magnesium, and to have an atomic weight 9-1. Other methods and considerations are re- sorted to in special cases, but these are more appropriately set forth at length in the usual textbooks of chemistry. All that need be noted specially in this place is the funda- mental fact that the numbers now recog- nised by chemists as representing the rela- tive values of atomic weights have been calculated upon the same scale and are all 23 THE ELEMENTS adjusted to a common standard. As already stated, a period exceeding half a century elapsed from the introduction and adoption of the Atomic Theory into chemistry before this uniformity was secured and before the relations of the elements to one another and to a general comprehensive scheme was recog- nised. It is out of a study of this scheme that modern speculations as to the nature and origin of all matter have chiefly origin- ated. CHAPTER [II INTERRELATIONS AMONG ATOMIC WEIGHTS "WE think in relations." HERBERT SPENCER. First Principles^ vi (ed. 1900), p. 145. FROM the previous chapter it is obvious that until the atomic weights had been reduced by common consent to one uniform scale or standard, it was not possible to per- ceive any general law governing the whole. Nevertheless many attempts were made to discover relations among the numbers in use. The first and one of the most famous of these is known in chemical literature as " Prout's hypothesis."* This assumes that the atomic * Ann. Phil., vi (1815), 321 j vii (1816), in. Prout's views were published anonymously under the title, On the Relations between the Specific Gravities of Bodies in their Gaseous State and the Weights of their Atoms. His thesis is summed up in the following passage (loc. cit. vii, 113): "There is an advantage in considering the volume of hydrogen equal to the atom, as, in this case, the specific gravities of most, or perhaps all, elementary sub- stances (hydrogen being i) will either exactly coincide with or be some multiple of the weights of their atoms." 25 THE ELEMENTS weights of all the other elements are multiples of the atomic weight of hydrogen. The case of chlorine, of which the atomic weight has always been known to be approxi- mately 35|, proved an insuperable obstacle to the adoption of Prout's original view, and subsequently attempts were made to fit the numbers which resulted from more and more accurate experiment to a unit assumed to to be first one-half, and subsequently one- fourth, of the atomic weight of hydrogen. Prout's hypothesis has been revived from time to time, but the progress of research has shown that in its original form, at any rate, the hypothesis is based on an illusion. But though a comprehensive scheme was not possible the recognition of families of closely allied elements led to much study of the numerical relations among their atomic weights. Among the elements first to attract notice were the halogens chlorine, bromine, and iodine ; the sulphur group sulphur, selenion, and tellurium; and the 26 ATOMIC WEIGHTS alkali metals potassium, sodium, and lith- ium. The nature of the relation observed will be sufficiently indicated by one case. Adopting the atomic weights that are used by Doebereiner in his memoir* for sulphur 32-239, selenion 79-263, and tellur- ium 129-243, it can be shown that these conform very nearly to an arithmetical pro- gression, for In other words, the atomic weight of selenion, which in properties stands between sulphur and tellurium, is very nearly the arithmetic mean of the atomic weights of the other two elements. The next step in advance is represented by the various attempts to establish an analogy between series of related elements and groups of carbon compounds, which about 1848 began to be arranged in " homo- * " Versuche zu einer Gruppirung der elementaren Stoffe nach ihrer Analogic." Pogg, Ann. (1829) !5> 3 01 - 27 THE ELEMENTS logous " series. Such a series consists of compounds of the same type containing, in addition to carbon, the same elements, ex- hibiting the same chemical characteristics, but differing from one another in mole- cular weight, and hence in physical properties. The composition of such a series is ex- pressible by a general formula in which, passing from any one term to the next above or below, there is a uniform difference of one atom of carbon and two atoms of hydrogen, or CH 2 . An example will render this quite clear. The normal primary alcohols. Name. Formula Molecular Boiling n 2n+2 weight. point. Methylic . CH 4 32 66 Ethylic . C 2 H 6 46 78-3 Propylic C 3 H S 60 97-4 Butylic . C 4 H 10 74 116-8 Amylic C 5 H 2 88 137 Hexylic . C 6 H 14 102 J57 Heptylic C 7 H 16 116 i?5 etc. etc. etc. etc. 28 ATOMIC WEIGHTS These are all compounds containing one atom of oxygen in the form of hydroxyl, HO, while the carbon and hydrogen increase by successive additions of CH 2 , and this in- crease is accompanied by a fairly steady rise in the boiling point, increase of specific gravity, and gradual change in viscosity till, in the highest terms not given in the table, the alcohol becomes a crystalline solid at common temperatures. Selecting any three successive terms the molecular weights show the same kind of relation, though with greater numerical ex- actitude, as that which is observed among the atomic weights in a natural family of elements. For example : Amylic+Heptylic Alcohol, 88+n6 = R .. A] 2 2. This is parallel with the relation existing among the alkali metals : Lithium + Potassium^; + 39'i -Sodium 23-0 2 2 29 THE ELEMENTS The formula of the alcohols may be written H 2 O-1-MCH 2 , and by giving to n any desired value the formula of any term of the series may be arrived at. Similarly the atomic weights of the ele- ments composing a natural family may be calculated by adopting the atomic weight of the first term as the basis to which addi- tions may be made of a common increment. Thus the alkali metals may be treated as derived from the basis 7, with differences equal to ni6, so that a being the value of the basis, a-\-nd gives the atomic weight, a = 7 = Lithium a+ d = 7 + 16 = 23 = Sodium a + 2d = 7 + 32 = 39 = Potassium In some cases it is necessary to assume two values for the increment, as in the case of the halogens and several other groups. Thus for the halogens : a = 1 9 Fluorine a+ d= 19 + 16-5 = 35-5 = Chlorine a + id+ d' = i9 + 33+28 = 80 = Bromine 38 + 33 + 56 = 127 = Iodine 30 ATOMIC WEIGHTS Several attempts in this direction were made in the middle of the nineteenth century, and of these the memoir of Dumas* repre- sents perhaps the most determined and comprehensive. It is unquestionably true that something akin to homology is to be traced in many families of elements, but at present in no case has a satisfactory formula been dis- covered for the calculation of the atomic weights so as to bring them into harmony with the values deduced from the most exact experiment s.f The arrangement of the elements in groups consisting of closely allied members does not, however, provide the clue to a scheme by which the whole of the elements could be shown to belong to one system of things. Prout's hypothesis seems to have been the only attempt to provide a general * ' ' Memoire sur les I^quiv. des Corps Simples. " Ann. Chim. Phys. [3], iv, 129. t See, however, later, chapter iv. 31 THE ELEMENTS law up to a period which commences about 1860-62. As already mentioned, this is attributable to two causes, namely, first the want of knowledge about the chemical characters of a considerable number of the elements ; and secondly, the want of co- ordination among the atomic weights. Of the chemists to whom the study of this question appealed strongly Odling should first be mentioned. Having occupied himself with the relations traceable among atomic weights from 1857 onward, Odling published, in 1864, an article containing a table in which the atomic weights of forty- five of the best-known elements are arranged horizontally in the order of their generally received groups, and perpendicularly in the order of their several atomic weights (W aits' s Diet., iii, 975). ATOMIC WEIGHTS Mo 96 Pd 106-5 W 184 Au 196-5 Pt 197 Li 7 9 B ii C 12 N 14 O 16 F 19 Na 23 Mg 24 Al 27-5 Si 28 P 31 s 32 Cl 35-5 Zn 65 As 75 Se 79-5 Br 80 Ag 108 Cd 112 Sn 118 Sb 122 Te 129 I 127 Hg 200 Tl 203 Pb 207 Bi 210 K 39 Ca 40 Ti 48 Cr 52-5 Mn 55 etc. Rb 85 Sr 87-5 Zr 895 Cs 133 Ba 137 V 138 Th 231 Here chromium and manganese are sup- posed to represent the metals of the iron group, and palladium and platinum their respective congeners, so that the total num- ber of elements provided for amounts to about fifty-two. It will be seen later that this table embodies a scheme which is only one step removed from the presentation of the periodic law which has since become the commonly accepted basis of classification. In the meantime other chemists and physicists were at work on the problem, and D 33 THE ELEMENTS in order that as far as possible justice may be done to the several publications which en- sued, mention must be made of some of them. In 1862 Beguyer de Chancourtois,* a French geologist, conceived the idea of repre- senting geometrically the relations among the atomic weights of the then known ele- ments by the device of a spiral drawn upon the surface of a cylinder with circular base divided into sixteen equal parts, the spiral cutting the generatrices (vertical lines) of the cylinder at an angle of 45. By measuring off from the base lengths corresponding to what he called " the characteristic numbers/' that is the atomic weights, de Chancourtois showed that elements of similar character often found places on the same generating line. Thus oxygen, sulphur, selenion, and tellurium stand vertically above one another, as also do the members of several other natural families, though intermingled wit{\ others not related to them, * Comptes Rendus, 1862 and 1863, 34 ATOMIC WEIGHTS But the language of the memoir is so obscure that it is uncertain whether the author really recognised the general relation- ship of properties to atomic weight.* At one time he seems to be discussing the distribu- tion of the elements in minerals, and at another to be confused by the fact that the atomic weights are not whole numbers, and the differences between them are not con- stant, and therefore cannot exhibit simple geometrical relationships on his " Vis tel- lurique," or telluric helix. Among other early attempts to trace a law of general application the several papers * The following passage extracted from one of his memoirs sufficiently explains his position : " La proposition fondamentale de mon systeme : Les rapports des proprittts des corps sont manifesto's par des rapports simples de position de leurs points caracttristiqucs. Par exemple, 1'oxygene, le soufre, le selenium, le tellure, le bismuth, s'alignent sensiblement sur une generatrice, tandis que le magnesium, le calcium, le fer, le strontium, 1'urane, le barium s'alignent sur une generatrice opposee ; a c6te de la premiere figurent d'une part 1'hydrogene et le zinc, d'autre part le brome et 1'iode, le cuivre et le plomb ; a cote de la deuxieme s'alignent le lithium, le sodium, le potassium et le manganese, etc. etc." Comptes RenditSi 54 (1862), 75$. 35 THE ELEMENTS of J. A. R. Newlands claim the most atten- tion, for this chemist was the first to draw attention definitely to the periodic character of the relation which is observable among the majority of the elements when a list is drawn up in the order of the numerical value of their atomic weights. Newlands' first paper (Chem. News, Feb., 1863), and several succeeding papers, treat of the rela- tions among the equivalents, and as at this time the relation of the atomic weights to the equivalents was still undetermined by any uniform rule, the research was not al- ways successful. In a communication made to the Chemical Society, in March, 1866, the author adopted Cannizzaro's system, and was thereby led to revise some of his previous statements, and succeeded in enunciating on a more satis- factory basis the " Law of Octaves " which he had formulated two years previously (Aug., 1864). His own words were as follows : ''If the 36 ATOMIC WEIGHTS elements are arranged in the order of their equivalents, calling hydrogen i, lithium 2, glucinum 3, boron 4, and so on (a separate number being attached to each element having a distinct equivalent of its own, and where two elements happen to have the same equivalent, both being designated by the same number) it will be observed that elements having consecutive numbers fre- quently either belong to the same group or occupy similar positions in different groups." And referring to a tabular arrangement of some of the elements he proceeded : " Here the difference between the number of the lowest member of a group and that immediately above it is 7 ; in other words, the eighth element starting from a given one is a kind of repetition of the first, like the eighth note of an octave in music." The following table was given later, when corrected atomic weights had been adopted. 37 THE ELEMENTS ELEMENTS ARRANGED IN OCTAVES No. H 1 Li 2 G 3 Bo 4 C 5 N 6 7 No. No. F 8C1 15 Na 9K 16 Mg lOiCa 17 Al HCr IS Si 12Ti 19 P 13 Mn 20 S 14 Fe 21 No. Co & Ni 22 Cu 23 Zn 24 Y 25 In 26 As 27 Se 28 No. Br 29 Rb 30 Sr 31 Ce & La 32 Zr S3 Di & Mo 34 Ro&Ru35 No. No. Pel 36ll 42 Ag 37lCs 44 Cd38JBa& V 45 U 39 Ta 46 Sn 40 W 47 Sb41|Nb 48 Te43IAu 49 No. Pt & Ir 50 Os 51 Hg 52 Tl 53 Pb 54 Bi 55 Th 56 (Chem. News, March 9, 1866.) At the meeting of the Chemical Society, when this table was brought forward, the objection was raised by Dr. Gladstone that no provision is made for elements which re- mained to be discovered. As the previous few years had brought forth caesium, rubi- dium, thallium, and indium there was con- siderable force in the remark, and to this curious mistake on the part of the author may be partly attributed the fact that he was not more successful in sustaining his fundamental idea. Nevertheless it must be admitted that Newlands was the first to announce definitely the discovery that the properties of the elements are a periodic function of their atomic weights, and this is clearly stated in his enunciation of his law 38 ATOMIC WEIGHTS of octaves, and is obvious from the positions assigned to a majority of the elements in the table. Unfortunately, he appears to have appreciated very imperfectly the importance of the principle involved, and he wrote no more on the subject till 1872, after the publication of MendeleefFs famous memoir, of which an account will now be given. From what has gone before it is obvious that the way was being prepared for a generalisation concerning the relation of properties to atomic weight, and it was therefore natural that Professor Mendeleeff, being occupied in the compilation of his well-known work entitled, in the English translation, The Principles of Chemistry, should have his attention strongly attracted to the subject. In March, 1869, Mendeleeff communicated to the Russian Chemical Society a memoir, of which an abstract only appeared, in German (Zeitsch. /. Chem., v, 405), and of which the following is a translation, slightly 39 THE ELEMENTS abbreviated, several obvious misprints being corrected, as well as one important error of translation, the Russian word for periodic having been rendered by the German " stufenweise," or gradual. " When the elements are arranged in vertical columns, according to increasing atomic weight, so that the horizontal lines contain analogous elements, again according to increasing atomic weight, the following arrangement results, from which several general conclusions may be derived : li- 50 Zr = 90 ? =180 V = 51 Nb 94 Ta =182 Cr= 52 Mo= 96 W =186 Mn= 55 Rh=104'4 Pt =197 '4 Fe= 56 -Ru =104'4 Ir =198 Ni=Co= 59 Pd 106-6 Os =199 H =1 Cu= 63-4 Ag =108 Hg = 200 Be- 9-4 Mg=24 Zn= 65-2 Cd =112 B =11 Al =27'4 ? = 68 Ur =116 Au=197? C =12 Si =28 ? - 70 Sn =118 N ~14 P =81 As- 75 Sb =122 Bi =210? O =16 S =32 Se= 79-4 Te =128? F =19 Cl =35-5 Br= 80 I =127 Li = 7 Na=23 K =39 Rb= 85-4 Cs =133 Tl =204 Ca-40 Sr* 87-6 Ba =137 Pb =207 -45 Ce= 92 r=56 La= 94 Yt=60 Di 95 ? In =75-6 Th = 118 \J ATOMIC WEIGHTS " i. The elements according to the mag- nitude of atomic weight show a periodic change of properties. "2. Chemically analogous elements have atomic weights either in agreement (Pt,Ir,Os), or increasing by equal amounts (K, Rb, Cs). "3. The arrangement according to atomic weights corresponds with the valency of the elements, and to a certain extent the differ- ence in chemical behaviour, for example, Li, Be, B, C, N, O, F. " 4. The elements most widely distributed in nature have small atomic weights, and all such elements are distinguished by their characteristic behaviour. They are thus typical elements and the lightest element, hydrogen, is therefore rightly chosen as the typical unit of mass. "5. The magnitude of the atomic weight determines the properties of the element. Hence the compounds of S and Te, of Cl and I show, beside many analogies, striking differences. 41 THE ELEMENTS " 6. It allows the discovery of many new elements to be foreseen; for example, analogues of Si and Al with atomic weights between 65 and 75. " 7. Some atomic weights will experience correction ; for example, Te cannot have the atomic weight 128, but 123 to 126." From the foregoing table the principle of periodicity, that is, recurrence of similar properties at regular intervals with increase of atomic weight, is less obvious than it afterwards became when, in 1871, the arrangement was modified so as to assume the form now commonly adopted in text- books of chemistry. In the meantime the study of the ques- tion led the German Professor Lothar Meyer to conclusions practically identical with those of Mendeleeff, and by his famous dia- gram of atomic volumes to illustrate very clearly the periodic recurrence of many of the physical properties of the elements when arranged in the order of atomic weights. 42 ATOMIC WEIGHTS It is only necessary to remind the reader that by " atomic volume " is meant the quotient which results from dividing the atomic weight by the density. It therefore represents the volume in cubic centimetres which would be occupied by the atomic weight of the element taken in grams. Meyer's table,* of which a portion is here reproduced, speaks for itself. Tracing al- most any well recognised natural family of elements, it can be seen from the curve that the successive terms occupy corresponding positions. For example, lithium, sodium, potassium, rubidium, and caesium occupy the successive apices and have the greatest known atomic volumes. This character is associated with great chemical activity. In like manner calcium, strontium, and barium are to be found in corresponding positions on the descending portions of the curve, while chlorine, bromine, and iodine, sulphur, selenion, and tellurium occupy places on the * Annalen (Dec., 1869). Supplement, p. 354. 43 \ X 1 ATOMIC WEIGHTS ascending portions. Other properties and relations are similarly recognisable. Nevertheless, Mendeleeff is rightly re- garded as the discoverer of the law. His title to be so regarded is based on the fact that he not only proclaimed the law afresh, and apparently in ignorance of Newlands' work, but he displayed that deep conviction of its important consequences which led him to some of the most striking and interesting of its applications. A detailed account of many of these is to be found in all the best text- books of chemistry ; it is unnecessary, there- fore, to do more than to indicate by a single example the application of the principle, which is embodied in Mendeleeffs revised table, to the prediction of new elements and a prevision of their physical and chemical properties. The table, in its latest form, as arranged by Mendeleeff, in 1904, to include all the known elements except the " rare earths/* is shown below. 45 THE ELEMENTS Series. .. Zero group. JK Group I. Group II. Group III. Group IV. _ 1 .. y Hydrogen, H = 1-008 - - - 2 .. Helium, He = 4'0 Lithium, Li = 7 '03 Beryllium, Be=9'l Boron, B = 11'0 Carbon, C = 120 3 .. Neon, Ne = 19-9 Sodium, Na = 23-05 Magnesium, Mg = 24'l Aluminium, Al-27'0 Silicon, Si = 28-4 4 .. Argon, Ar=38 Potassium, K = 39'l Calcium, Ca=40'l Scandium, Sc = 44'l Titanium, Ti = 48'l 5 .. - Copper, Cu = 63-6 Zinc, Zn = 65'4 Gallium, Ga=70'0 Germanium, Ge=72'3 6 .. Krypton, Kr = 81-8 Rubidium, Rb = 85-4 Strontium. Sr=87'6 Yttrium, Y = 89'0 Zirconium, Zr=906 V - Silver, Ag = 107'9 Cadmium, Cd = 112-4 Indium, In = 114-0 Tin, ' Sn =119-0 8 .. Xenon, Xe = 128 Caesium, Cs = 132'9 Barium, Ba*137'4 Lanthanum, La = 139 Cerium, Ce = 140 . . 10 .. Ytterbium, Yb=173 11 .. - Gold, Au -=197-2 Mercury, Hg = 200-0 Thallium, Tl= 204-1 Lead, Pb = 206'9 12 :. '- - Radium, Rd = 224 - Thorium, Th = 232 46 ATOMIC WEIGHTS Group V. Group VI. Group VII. Group VIII. Nitrogen, N=14'04 Oxygen. O = 16-0 Fluorine, F-19-0 . Phosphorus Sulphur, 5 = 32-06 Chlorine, Cl-35'45 - - - Vanadium, V = 51-4 Chromium, Manganese, Mn = 55'0 Iron, Cobalt, Nickel, Fe = 55'9 Co = 59 Ni = 59 (Cu) Arsenic, As = 75-0 Selenium, Se=79'0 Bromine, Br-79-95 - - . - Niobium, Molybden'm Mo =96-0 - Ruthenium, Rhodium, Palladium. Ru = 101'7 Rh = 103 -0 Pd = 106-5 (Ag) Antimony, Sb = 120-0 Tellurium, Te = 127 Iodine, 1 = 127 Tantalum, Ta= 183-0 Tungsten, W = 184 Osmium, Indium, Platinum, Os = 191 Ir = 193 Pt = 194-9 (Au) Bismuth,', Bi = 208 - _ __. _ - Uranium, U=239 :. _ --''--. - 47 THE ELEMENTS In 1871, when Mendeleeff drew up this table, the spaces corresponding to atomic weights 44, 70, and 72 were vacant. To the hypothetical elements expected to occupy these positions he gave the names ekaboron, akaluminium, and ekasilicon, and to each he assigned the properties which were soon afterwards recognised in the new elements scandium (at. wt. 44), gallium (at. wt. 69-9), and germanium (at. wt. 72-5). To the element standing next after zinc Mendeleeff gave the name eka-aluminium, and the following is the outline he gave of its properties :* "It will be in the same group as Al, and should consequently give R 2 O 3 , RC1 3 , R 2 (SO 4 ) 3 , alums, and like compounds analogous to those of aluminium. Its oxide should be more easily reducible to metal than alumina, just as zinc oxide is more easily reduced than magnesia. The oxide R 2 O 3 should, like alumina, have feeble but clearly expressed basic properties. The * Principles^ ii, 84 (Engl. ed., 1891). 48 ATOMIC WEIGHTS metal reduced from its compounds should have a greater atomic volume than zinc, because in the fifth series, proceeding from zinc to bromine, the volume increases. And as the volume of zinc is 9-2, and of arsenic 18, therefore that of our metal should be near to 12. This is also evident from the fact that the volume of aluminium =11, and of indium = 14, and our metal is situated in the III group, between aluminium and indium. If its volume is 11-5 and its atomic weight be about 69, then its density will be nearly 5-9. The fact of zinc being more volatile than magnesium gives reason for thinking that the metal in question will be more volatile than aluminium, and therefore for expecting its discovery by the aid of the spectroscope, etc." In 1875 Lecoq de Boisbaudran discovered by means of the spectroscope a new metal in a zinc blende from the Pyrenees. This he named gallium, and it was found by subse- quent study to have the atomic weight 69-8, E 49 THE ELEMENTS the density 5-9, to form a sesquioxide Ga 2 O 3 , and an octahedral alum, like common alum. The metal is soluble in acids and in alkaline hydroxide, and possesses many of the pro- perties of aluminium. It is, however, much more fusible, melting at 30 ; just as zinc is more fusible than magnesium. Similar predictions concerning the other two hypothetical elements mentioned above were completely confirmed by the properties observed in the metals scandium and ger- manium. No justification could be more complete, and Mendeleeff's scheme has con- tinued to furnish the guiding principle of the greater part of modern inorganic chemical research. CHAPTER III THE PERIODIC LAW "As in Mathematics so in Natural Philosophy, the investi- gation of difficult things by the method of analysis ought ever to precede the method of composition." NEWTON, Optic ks. HAVING traced the gradual course of development of the idea which is em- bodied in Mendeleeff's scheme of the elements it will now be useful to examine more closely his statement of the " Periodic Law/' Mendeleeff 's own words rendered into Eng- lish in the latest edition of his Principles (1905, vol. ii, p. 17) appear as follows : " The properties of the elements, as well as the forms and properties of their compounds, are in periodic dependence on, or (express- ing ourselves algebraically) form a periodic function of, the atomic weights of the elements." In order that the question of the general THE ELEMENTS validity of this law may be fairly considered it is necessary first of all to inquire whether there is any reason to expect the discovery, by further research, of substances of ele- mental character at present unrecognised. In order to answer this question the first thing to do is to examine the list of elements now generally acknowledged, and it will be found that they form a continuous series, with a roughly uniform progression in the value of the atomic weight in passing from term to term. The following table contains the names and atomic weights of the eighty- one elements recognised by the International Committee on Atomic Weights with the numerical values adopted by that body. ELEMENTS ARRANGED IN ORDER ACCORDING TO ATOMIC WEIGHT Hydrogen . . 1-008 Helium . . . 4-0 Lithium . . 7-00 Beryllium (Glucinum) 9-1 Boron . . il-o Carbon . . . 12-00 Nitrogen Oxygen . Fluorine Neon . Sodium . Magnesium . 14-01 . 1 6-00 . 19-0 . 2O-O . 23-00 . 24-32 52 THE PERIODIC LAW Aluminium . . 27-1 Indium . 114-8 Silicon . . 28-3 Tin 119-0 Phosphorus . . 31-0 Antimony 120-2 Sulphur . 32-07 *Tellurium 127-5 Chlorine Iodine . I26-92 *Argon . 39-9 Xenon . I30-7 Potassium . 39-10 Caesium I32.8I Calcium . 40-09 Barium . 137-37 Scandium . 44-1 Lanthanum 139-0 Titanium . 48-1 Cerium . I40-25 Vanadium . 51-2 Praseodymiur n I40-6 Chromium . 52-0 Neodymium 144-3 Manganese . 54-93 Samarium I50-4 Iron Nickel 55-85 . 58-68 Europium Gadolinium 152-0 157-3 Cobalt 58-97 Terbium I59-2 Copper Zinc 63-57 65-37 Dysprosium Erbium . 162-5 167-4 Gallium . 69-9 Thulium 168-5 Germanium . 72-5 Ytterbium 1720 Arsenic . Lutecium 1740 Selenion . 79-2 Tantalum 181-0 Bromine . 79-92 Tungsten 184-0 Krypton . 83-0 Osmium 1909 Rubidium . 85-45 Iridium . I93' 1 Strontium . . 87-62 Platinum 195-0 Yttrium . 89-0 Gold . 197-2 Zirconium . 90-6 Mercury 200-0 Niobium(Columbium)93'5 Thallium 2O4O Molybdenum . . 96-0 Lead . 207-IO Ruthenium . . 101-7 Bismuth 208-0 Rhodium . 102-9 Radium 226-4 Palladium . 106-7 Thorium 232-42 Silver . . 107-88 Uranium Cadmium . 112-40 (Total 8 1.) * Argon and tellurium are placed out of numerical order on account of the uncertainty still attaching to the relative values of their atomic weights and those of the elements immediately following them. 53 THE ELEMENTS It will be observed that in passing from one element to the next in the list the differ- ences between the atomic weights vary from 0-29 (Co - Ni) to 4-6 (Cu - Co), among the ele- ments of which the atomic weights are not very large and most of which have been determined with considerable approach to accuracy. The relatively large difference, 7-3, be- tween antimony and tellurium is attributed to some error in the atomic weight of tellurium, of which no sufficient explanation has yet been found ; but there is a gap be- tween molybdenum and ruthenium amount- ing to 57 units, which is supposed to indicate a vacancy appropriated in the Mendeleeff scheme to a homologue of manganese. The atomic weights of the fourteen elements, beginning with lanthanum, are confessedly uncertain, but that they all lie between lanthanum and tantalum seems probable, because, although the individual numbers are doubtless inexact, the average difference 54 THE PERIODIC LAW between any two consecutive terms is ap- proximately the same as the average differ- ence between successive atomic weights among the better known elements preceding them. Ta - La=i8i 139=42 for thirteen intervals, or about 3-2. Between tungsten and osmium a differ- ence of 6-9 units seems to indicate something missing, and from bismuth, 208, to radium, 226-4, there is a wide interval which seems to indicate about four vacant places. It is, of course, uncertain whether the same order of increase is to be ex- pected in the larger numbers, and whether some irregularity is possible or even prob- able here. It should also be noted that the differences, approximately three units each, among the three elements with smallest known atomic weights, namely : H i -008 He 4-0 Li 7-00 are greater than the differences observed 55 THE ELEMENTS among the elements immediately following them, namely : Li 7-00 Be 9-1 BII-O C 12-00 N 14-01 O 16-00 F 19-0 Ne 20-0 which show an average difference of less than two units between successive terms. At present no element is known with a smaller atomic weight than hydrogen or a larger atomic weight than uranium. Re- membering that the " atomic weights " are only ratios and represent only the relative magnitude of the masses of the atoms, and not their absolute masses in terms of any standard, there is nothing in theory to pre- clude the expectation of additions of new substances to either extremity of the series. However, the spectroscopic simplicity of hydrogen seems to hint that its constitution is near the limit at one end, and the belief that radio-activity is occasioned by the in- stability of the larger atoms at the other end leads to the suspicion, if not the convic- 56 THE PERIODIC LAW tion, that the series is limited by this in- stability, which, so far as is at present known, is associated with an atomic weight approximating to 240. From this point of view, then, the total number of new elements to be expected is not large, and they are for the most part such as would exhibit metallic characters and a high atomic weight. The significance of these differences among the successive atomic weights is more easily recognised when the elements are arranged in such a table as that of Mendele'eff (p. 46), where their natural affinities are brought into view and a number of natural families can be at once selected, each possessing well- marked characters common to the whole family. The most strongly marked of these families are the following : 1. Helium, neon, argon, krypton, xenon. 2. Lithium, sodium, potassium, rubidium, caesium. 57 THE ELEMENTS 3. Beryllium, magnesium, zinc, cadmium, mercury. 4. Calcium, strontium, barium, radium. 5. Aluminium, gallium, indium. 6. Silicon, titanium, zirconium. 7. Germanium, tin, lead. 8. Nitrogen, phosphorus, arsenic, anti- mony, bismuth. 9. Vanadium, niobium, tantalum. 10. Oxygen, sulphur, selenion, tellurium. 11. Fluorine, chlorine^ bromine, iodine. 12. Chromium, manganese, cobalt, iron, nickel, copper. 13. Chromium, molybdenum, tungsten, and uranium. 14. Ruthenium, rhodium, palladium, os- mium, iridium, platinum. Here it may be to the advantage of the reader if a sketch is given of the characters by which one or two of these families are distinguished, in order to indicate the sort of features which are regarded as important in determining relationships. These are 58 THE PERIODIC LAW at once divisible into two kinds, namely, (i) those which are recognised as matters of fact, such as the physical and chemical pro- perties of the elements and of their chief compounds ; and (2) valency, which is to some extent involved in theory. Concerning the former it will be found in all cases that density, fusibility, or volatility, and chemical activity are obviously related to atomic weight, the density always in- creasing with increasing atomic weight, while volatility generally diminishes among the non-metals, and increases among the metals with rise in the value of atomic weight. As to valency, br the capacity displayed by the atom of the element to associate with other atoms, ^ if used alone this character would bring together quite incongruous materials, while it sometimes separates very similar substances. Phosphorus and sulphur, for example, are very much alike, as are also antimony and tellurium and other elements, which are properly placed in separate 59 THE ELEMENTS families in consideration of their differences of valency. The alkali metals may be taken as an example. These substances are silvery white, very fusible, volatile solids, which communicate to flame characteristic colours, those given by potassium, rubidium, and caesium being purplish and almost indistin- guishable from each other by the eye. By means of the spectroscope these lights are resolvable into a very small number of bright lines. These metals decompose water violently and are distinguished by their tendency to unite with oxygen, their electro- positive character increasing as the atomic weight increases. Their salts are with very few exceptions easily soluble in water, and the corresponding salts are usually isomor- phous. Thus the chlorides, bromides, and iodides, when anhydrous, crystallise in cubes, and the sulphates, with the exception of lithium sulphate, unite with aluminium sulphate and water to form alums which 60 THE PERIODIC LAW crystallise in regular octahedrons. Their chief physical properties are shown in the following table : Li Na K Rb Cs Atomic Wt. . 7*0 23*0 39*1 85^45 132*81 Density . . 0-59 0*97 o'8; 1-52 r88 Atomic Vol. . ii'9 23-6 44-9 56*2 70*6 Melting Pt. . 186 95 62 38*5 26 Boiling Pt. above red 742 667 ? ? These elements are all univalent, that is, one atom of the metal is capable of combin- ing with or of displacing one atom of hydro- gen, and there is no well recognised evidence that they ever show a greater capacity for combination. Turning now to the opposite side of the table, the halogens may be taken as an ex- ample of a well-defined natural family. These are all very volatile substances, the vapours of which exhibit characteristic colours ; fluorine pale yellow, chlorine green- ish yellow, bromine orange-red, iodine deep purple. By reason of their low boiling points fluorine and chlorine are gaseous at atmo- spheric pressure and temperature, while bromine is a red dense liquid, and iodine a 61 THE ELEMENTS black, lustrous, crystalline solid. They are distinguished by the tendency they exhibit to combine with hydrogen and metals, and not, by any direct process, with oxygen. Fluorine is the most electro-negative element known, displacing chlorine from the chlorides and entering into combination with hydro- gen even in the dark and at very low tem- peratures. It seems to be incapable of com- bining with oxygen under any circumstances. The following table exhibits the chief physical properties of these elements, and as in the case of the metals it may be noticed that these properties follow the numerical magnitude of the atomic weight, the chemi- cal activity standing in the inverse order, fluorine being the most active, while iodine is the least. F Cl Br I Atomic Wt. . . 19-0 35-46 79-92 126-92 Density gas (H=i) 19 35'5 80 127 ,, liquid I 14 (at -200) i '42 3'l8 ,, or solid - . 4 '9 Atomic Vol. . . . i6'6 25-0 2S'I 257 Melting Pt. . . below -210 7 to - 8 114 Boiling Pt. . at atmos. pressure . } -187 -33 63 184 62 THE PERIODIC LAW In this family, as in many others, the first and last members of the series exhibit abnormalities, but it is unnecessary in this place to enter into details regarding the peculiarities of fluorine, on the one hand, and of iodine on the other. The valency of the group in their com- pounds with hydrogen and the metals is habitually represented by one unit ; but in certain compounds, notably in those which contain oxygen, it is necessary to assume a higher degree of combining capa- city, amounting to three, or, according to some chemists, five or even seven, units. If the table of elements drawn up by Mendeleeff is now inspected it becomes at once obvious that, after Series i, of which hydrogen is the only acknowledged member, the two following lines illustrate perfectly the principle of periodicity. Here the ninth element, counting from the first, exhibits an almost perfect reproduction of the same assemblage of properties as the first with such modification as might be 63 THE ELEMENTS expected from the increase of density which follows the increase of atomic weight. The fourth series, however, shows greater differ- ences, first in respect to the positions assigned to chromium and manganese respectively, which in their metallic character are wholly unlike the sulphur and halogen groups to which they are attached ; and secondly, in the fact that the atomic weights of iron, nickel, and cobalt bring these elements into close association with manganese and to pro- vide places for them it is necessary to assume an extension of the period by the addition of an eighth group. Admitting this modifica- tion, the characteristics of these metals should be revived later on among elements of higher atomic weight. The platinum metals do in two series represent correlatives of the iron group which they resemble in difficult fusibility, in the tendency to occlude hydrogen and in the formation of complex cyanides and ammines. Complete revival of characteristics is not to be looked for, and there is considerable modification of valency. 64 THE PERIODIC LAW Passing over the difficulties encountered in respect to individual elements such as those already mentioned and others, the most serious problem is presented by the elements related to cerium, and constituting the substances long known as the " rare earths." Thirteen of these substances are recognised by the Atomic Weight Committee, and will be found in the table of atomic weights, commencing with lanthanum, 139, and ending with lutecium, 174. These metals form sesquioxides, and their sulphates constitute an isomorphous group, with the general formula R 2 (SO 4 )3,8H 2 O. Some of them have been reduced to the metallic state and their densities have been found to approximate to the density of iron, but with much lower melting-point. Dens. M.P.* Cerium . . . 7-04 623 Lanthanum . .6-15 810 Praseodymium . . 6-47 94 Neodymium . . 6-96 840 Samarium . -7*75 (Iron . . . .7-84 above 1600) * Muthmann and Weiss, Annalen, 1904, 331, I. F 6 5 THE ELEMENTS Two kinds of difficulty arise in connection with the placing of these metals in the scheme. They are all usually trivalent, forming sesquioxides, and therefore cannot fall into successive groups with different valencies, and they are much less easily reducible than the platinum metals ; and in respect to density they do not fall between the two subdivisions of those metals, the lighter with density about 12, and the heavier with density 21-22, shown in Group VIII. Mendeleeff's enunciation of the periodic principle, therefore, can only be accepted with some qualification. An inspection of the table shows that following Series 3 there is an obvious division of each group vertically into two sub-groups, one of which preserves, substantially, the character of the type ; the other displays a very rapid development of quite new fea- tures. This is perhaps best shown by tabu- lating the former separately, as follows : 66 J5 ^. ON ON a I r ^ I ? I 2 1 1 1 1 1 2 114 u w w O gj VO N ON t^ a HH CO 1 t- 1 2 O CO JJ 1 1 1 1 1 O g ON N ex Tj- hH fr O i 1-1 co i r^ i N 'Z P-t 1 w ' 1 1 OO 1 1 1 N 1 O CO PQ > a CO w vp M 00 CO O I ~ (N Tf I ON I M b - 1 ;fr ri 1 I ff O ^ co H N p 1 ? ' * \ k 1 f U ^ CO CO ! $ 1 1 tj O * 00 a M t^ co ON m n CO , i | | O a rt u ON ^ rj- o ON co O 2 | _ | HH J^J ^ k^ ? | u 1 1 1 X i t-i N to TJ- m vo r^ 00 ON O M M i THE ELEMENTS Here it will be perceived that as atomic weight increases in each vertical column there is a tendency to increase in the electro- positive character of the element, indicated among the metals by the greater difficulty encountered in the isolation of the metal from its compounds and in the more strongly marked basic character of the oxides. Among the non-metals the increase of atomic weight in any group is attended by gradual development of metalloidal characters, as may be noticed in passing from phosphorus to antimony, from sulphur to tellurium, and from chlorine to iodine. The elements omitted are all metallic in character, and, as with the rest, their metallic qualities in each group suffer modification, but in the inverse order, for reduction to the metallic state is here generally easier among the elements of highest atomic weight, and the basic properties of the oxides are most strongly marked among those of lower atomic weight. This may be recognised in 68 THE PERIODIC LAW following the successive members of such a family as copper, silver, gold, or zinc, cadmium, mercury ; or by comparison of the iron group with the two sub-groups of platinum metals. The elements referred to are arranged in their several groups and series in the follow- ing table. (See next page.) 69 I I <3| I g I I I fig I i I O ' I I I * I pTg I I I ~ I I I fi I Ss I I I 58. I i i i ie I I I I I I I i i i>isg 2 I I I I I I o I -J I I I H| SM Ssrl i i K 8 i I I I OS !. 25 H U H N c5 S CQ i o rt J3 CL ' < W2 ^ ^ ^ . bfl M CO M PQ ,s a a THEORIES OF EVOLUTION Thus the direct process in the first three columns is easily intelligible, but to account for the ten elements following titanium it is suggested that they are the results of some indirect process of evolution from silicon. It may be observed that the direct process leads to the formation of elements of precisely the same general character and identical valency, while the indirect process leads to the formation of elements of different valencies. The authors also make use of the idea of possible devolution, and regarding iron, nickel, and cobalt as abnormally constituted derivatives of manganese, they consider copper, zinc, gallium, and germanium as the products of a return to more normal structure. Silver, cadmium, indium, and tin are similarly regarded as recovery products of ruthenium, rhodium, and palladium ; while gold, mercury, thallium, and lead are successive recovery products from osmium, iridium, and platinum. There is much ingenious argument in this 87 THE ELEMENTS paper, and the superior probability of direct over indirect evolution of the natural families so far as this principle is applied in the paper is a point of interest. But the assumption of the four protons is a source of difficulty, and as will be seen later is a less probable hypothesis than that of two. The authors admit the anomalies in the atomic weights of tellurium and argon, which according to all available experimental evidence are greater than those of iodine and potassium respectively. This admission, it must be remembered, however, is not in harmony with Mendeleeff's periodic law. The arrange- ment of the elements in the table will not commend itself to all chemists on account of the disruption of familiar associations, such as the separation of zinc from mag- nesium, vanadium from phosphorus, tin from the carbon group, and so forth. In the Mendeleeff chart of the elements nothing is more striking than the gathering together of the negative elements into one 88 THEORIES OF EVOLUTION corner of the table, fluorine, the most electro- negative element known, being at the ex- tremity opposite to the position of lithium at the head of the most electro-positive of the metals, which are flanked by the zero group of inactive elements, helium, argon, etc. If the table were wrapped round a cylinder, the groups being vertical, this argon group would stand between positive and negative. It is noticeable that the elements exhibiting more or less well-defined metallic characters are far more numerous than those which present negative charac- ters, and the physical properties of the latter are far more diverse. There is an unmis- takable family likeness about the substances called metals which, coupled with their rela- tively large number, seems to indicate that this form of matter was more easily produced or was relatively more stable under the conditions of high temperature or electrical stress which probably characterised the initial stages of evolution. 89 THE ELEMENTS The metals are, as a rule, solid, lustrous bodies possessing relatively high conductivity for heat and electricity, together with malleability and ductility. Mercury, which is so fusible as to be liquid at common tem- peratures, possesses the above-mentioned characters when in the solid state. On the other hand, there are, of course, several which are deficient to some extent in one or other of the common characteristics. They all agree, however, in being deposited at the cathode when any of their compounds are submitted in the liquid state to the action of an electric current. The non-metals comprise a miscellaneous set of substances of which the majority are distinguished by being liberated at the anode in the course of electrolysis. Several of them, however, especially carbon, are not known to be deposited electrolytically. As to their physical condition at common tem- peratures, some are almost infusible solids, like carbon and silicon ; some are easily 90 THEORIES OF EVOLUTION melted, like phosphorus and sulphur ; one is a liquid bromine and several are gases, like nitrogen, oxygen, fluorine, chlorine. Roughly dividing the chemically active elements into the two classes, and omitting hydrogen, we find there are fifty-two metals to thirteen non-metals. Now if the members of the several groups from I to VII be con- sidered, reading vertically downwards, it will be found that as the atomic weight in- creases there is generally a tendency to an increase in the positive characters of the element ; but among the negative elements there is never an increase in the negative character. Thus there is a well-marked in- tensification of positive character in pass- ing down the group potassium, rubidium, caesium, and there is something very similar observable in the next group, calcium, stron- tium, barium. Sodium, copper, silver, and magnesium, zinc, cadmium, do not show a corresponding development of the positive character ; but in these cases there are 91 THE ELEMENTS other anomalies such as, in the latter group, fusibility and volatility increasing with atomic weight instead of the reverse as usual. On the other hand, chlorine, bromine, iodine, and sulphur, selenion, tellurium or phosphorus, arsenic, antimony, exhibit in each group a tendency to the suppression of the negative or chlorous character, and even to the development of metallic appear- ance and basic properties, as the atomic weight is increased. Berzelius, a century ago, attempted to explain the relative positions of the elements in an electro-chemical series by the assump- tion that each atom carries charges of positive and negative electricity, the pre- ponderance of one or the other serving to determine the chemical character of the sub- stance. In more recent times the idea of the co- existence of two elementary principles in the atoms of the chemical " elements " has been discussed more than once. 92 THEORIES OF EVOLUTION Carnelly, in 1885 (Brit. Assoc. Reports), brought forward the idea that these sub- stances are not strictly simple or elemental, but are compound radicals made up of, at least, two simple elements A and B. The element A was supposed to be identical with carbon, while B was connected with a negative weight, - 2, and it was suggested that it might be the ether of space. The conception of a negative weight has never been acceptable, and the hypothesis has for many years dropped out of sight. According to another suggestion by C. S. Palmer (Proc. Colorado Scient. Soc.),* the existence is assumed of two sub-elements, to which the names " kalidium " and " oxidium " were given. The hypothesis that hydrogen is the proximate ingredient of the elements was discarded by the author because the atomic weights have not been found to be exact * The original article is abstracted in Venable's Periodic Law, and is referred to in footnotes in Palmer's translation of Nernst's Theoretical Chemistry. 93 THE ELEMENTS multiples of unity, and because hydrogen is inherently basic, and while it might be looked upon as the prototype of base- forming elements, it could not be the origin of the acid-forming elements. The author suggests that possibly hydrogen is a member of a complete independent series as yet un- known. The properties of the last element of this series, or prefluorine, were discussed by Palmer, and here he seems to have antici- pated Mendeleeff. As to kalidium and oxidium, the two hypothetical components of all the elements, they are not regarded by Palmer as isolable forms of matter, but merely as representing antithetic qualities which are jointly responsible for the pro- perties of the elements as we know them. Within the last ten years investigations on the discharge of electricity through gases, especially by Sir J. J. Thomson and his school, have led to the development of a corpuscular theory of matter which, when more fully developed, will probably go far to 94 THEORIES OF EVOLUTION account for many of the chemical characters of the elements and their periodic relation to atomic weight. The following is a brief outline of the theory, of which details should be studied in The Corpuscular Theory of Matter, by J. J. Thomson.* When any gas, enclosed in a highly ex- hausted glass tube so that the gas pressure is exceedingly small, is exposed to an electric discharge, the rays proceeding from the cathode in straight lines exhibit several distinctive properties. They cause a phos- phorescence of the glass surface upon which they strike, and they are deflected by a magnet and by an electrified body in such a manner as to indicate that they consist of streams of negatively electrified particles. They also are capable of penetrating thin sheets of certain metals, and it becomes obvious that these particles are much smaller than the atoms or molecules of ordinary * Constable and Co., 1907. 95 THE ELEMENTS gases. By experimental methods, of which an account is given in the work cited, Sir J. J. Thomson has shown that these small particles or " corpuscles," as he called them, have a mass which is only about T^ of the mass of a hydrogen atom. Each corpuscle carries a charge of negative electricity equal to that which is carried by an atom of hydrogen in the process of electrolysis. The separation of a corpuscle from an atom im- plies that the residue retains an equal posi- tive charge. The carriers of positive elec- tricity are, however, not corpuscles like those which carry negative electricity, but seem to be masses of which the least is comparable with the mass of an atom of hydrogen. Negative corpuscles are producible not only from attenuated gases under the action of electric discharge, but are obtained from all kinds of matter, such as metals at a red heat, from heated oxides like lime, and from radio-active substances such as ura- nium and radium. 96 THEORIES OF EVOLUTION The corpuscular theory of matter assumes that an atom of any element consists of a mass of corpuscles, which, being all elec- trically negative, repel one another, and must therefore be held together by the presence of positive electricity equivalent in amount, so as to produce that electrical neutrality which is the condition of the atoms in their normal state. There seems to be no definite knowledge at present as to the form in which the positive electricity exists in an atom, but as already stated no positively electrified body has been found having a mass much less than that of an atom of hydrogen. The distribution of negative particles within a sphere of positive elec- tricity of uniform density has been the sub- ject of mathematical investigation by the author of the theory,* with results which are extremely interesting from the point of view of the chemist. When the presence of only one corpuscle * Phil. Mag., March, 1904, H 97 THE ELEMENTS is assumed it obviously goes to the centre oi the sphere. When a larger number are supposed to be present the cases in which they are confined to a plane passing through the centre of the sphere have alone been in- vestigated. It is shown in the paper referred to that five is the greatest number of corpuscles which can be in equilibrium in a single ring, but if others are placed within the ring then a larger number can maintain equilibrium in the ring. Thus, though a ring consisting of six corpuscles placed at the corners of a regu- lar hexagon is unstable by itself, it becomes stable when one corpuscle is placed in the centre of the hexagon. A greater number will arrange themselves in a series of concentric rings, the number of corpuscles in each ring increasing as the radius of the ring increases. The following table shows the numbers of corpuscles from one to sixty-nine, arranged in rings, the first row showing the numbers which may fall into one ring, the second^ 98 THEORIES OF EVOLUTION series the numbers which may produce two rings, the third series those which produce three rings, and so on. (See next page.) The numbers in the same vertical columns are repeated from series to series, the in- creased number of corpuscles in the addi- tional ring forming the top line. Thus, in the first column, we find in the first series 5, i ; in the second series n, 5, i ; in the third series 15, n, 5, i ; and in the fourth 17, 15, n, 5, i. This similarity of arrange- ment means, that supposing atoms so con- stituted they would have similar properties, and the substances so formed would present the characters of a natural family, such as are to be found in the " groups " of the periodic table of elements. The relation of these configurations to the periodic scheme of the elements may be further shown as follows : Consider the properties of all the configurations which have 20 corpuscles in the outer ring. (See following table.) The smallest number with 99 CD i t-H vo vo VO vo >0 vo s . r^ to i-i to o j^r HH * u ^ to o o O M Cr en * OS ^? * r^ T}- o to ^ en o 1 1 rn 1 to o to > ' H* t^ rf O ^ S - eo 0> en 8 ^f O vo 2 + g VO rn Ot to g 2" "* ^ en O CO s VO en OO en i-i vo en O -* t^ en ON en Q vo en CO M co h* t t d M * * | O to en O en \O en ON en ON vo en 00 M l-l - 2 * > <* ^ vo f^ oo en ON VO 1 M CO * 8 O en en 00 f> vo en 00 N ON VO M CO r ON en en OO CO VO N 00 N ON VO 2 ^ w O 00 m N 00 N VO N 00 M 00 t ( vo 2 K - 1 w ^ OO W W OO *^ VO N t^ M 00 VO M K M M ^ *"* *^ ""' M s . oo Ht N r>. w VO t^ +* 00 VO t - * en ^ ^ M vo = t. M 00 vo ^ * '-' n VO w M vo - to M vo M OO vo i-i * - M U1 M l- VO W to i-i to i-i ^ VO M vo M' THEORIES OF EVOLUTION an outer ring of 20 is 59, and in this case the number within is only just sufficient to render the ring of 20 stable, consequently any disturbance from without may cause the outer ring to shed one negative corpuscle, whereby the residue of 58 acquires a charge of positive electricity, and the atom so con- stituted would present the character of a univalent positive element if it could retain the charge. Passing from 59 to 60 the outer ring is more stable because there is one more cor- puscle within, and the stability goes on in- creasing till there is a total of 67 corpuscles in the atom. The addition of one more corpuscle makes the arrangement unstable, because it now goes into the outer ring, which then consists of 21 corpuscles, and one of these can be easily detached, as in the case of the arrangement of 59 corpuscles already considered, and the result is another electro- positive atom. Now the change from 59 to 67 corresponds 101 THE ELEMENTS to the addition of 8 corpuscles, but each addition of one corresponds to a new arrange- ment in the inner rings, as shown by the gradually increasing numbers in the lower rows of figures in the table, thus 16 to 17 in the second row, 13 to 14 in the third, 8 to 10 in the fourth, and 2 to 5 in the last row. So that as additional corpuscles are introduced the stability of the system in- creases till the total number is 67, while the addition of one more, just as the subtraction of one from 59, would produce instability. It can be shown, then, that a series of atoms constituted on this hypothesis, with 20 corpuscles in the outer ring, would possess valencies corresponding to those assigned to the elements in the first two series in the periodic scheme ; thus : Number . 59 60 61 62 6j 64 65 66 67 Valencv!' +O + 1 +2 + 3 + 4 -| -1 -I - Valency j _ g _ ; _ 6 _ s _ 4 +5 +6 +? +8 Total .. 888888888 The elements corresponding are : He Li Be B C N O F Ne Ne Na Mg Al Si P S Cl A 102 THEORIES OF EVOLUTION The constant sum of positive and negative valencies in the series quoted seems to sug- gest Abegg's hypothesis of normal and con- tra valencies. It must, however, be observed that the parallel between the hypothetical series shown above and the elements as they stand in the periodic scheme is, in the present position of the hypothesis, far from satis- factory, if only for the reason that there is a difference of only one corpuscle between the successive terms. The corpuscle has a mass equal to about TT Vu of the hydrogen atom, while the differences between the known elements He, Li, Be, etc., are much greater and are not uniform in passing from one element to the next. Moreover, as pointed out by Thomson himself, the agreement between the hypo- thetical and actual, shown in the numbers just quoted, is merely accidental, and until the mathematical difficulties of an investiga- tion in which the corpuscles are not confined 103 THE ELEMENTS to one plane have been overcome the theory cannot be further tested. The number of corpuscles in an atom is probably greater than the numbers assumed in the previous argument ; but on this point there seems to be no satisfactory evidence at present. There is, however, reason to think that whatever the number of corpuscles present the valency of the atom would vary periodi- cally with the number, that is, with the atomic weight. The word corpuscle which has been used throughout the previous statement, ab- stracted from Sir J. J. Thomson's book, is now usually replaced by the word " electron/' a term originally introduced by Dr. John- stone Stoney to designate the unit or atomic quantity of electricity. That this atomic quantity of electricity is separable from atoms of ordinary matter has been shown by work on the electric discharge through gases already referred to, and the idea that the electrical condition of matter, and its 104 THEORIES OF EVOLUTION chemical activity, whether positive or nega- tive, depend on the addition or removal of electrons has of late found much favour. Sir William Ramsay,* adopting this hypothesis, has suggested a notation for representing chemical combination, the link between two atoms being represented by the appropriate number of electrons represented by the symbol E. Common salt, for example, would be expressed on this system by the formula NaECl. This seems to imply that the electron or electrons which are supposed to be the cause of valency and to incite atoms to combine are external to the atoms concerned. This is a proposition which would admit of argument, and has indeed been challenged. The evi- dence available for either view is at present only of an indirect character, and is not ripe for discussion. Another way of treating the question is * Presidential Address to the Chemical Society, J. Chem. Soc., xciii, 774 (April, 1908). 105 THE ELEMENTS represented by the recent paper of Dr. James Moir.* The author assumes the cause of valency, at all events of the funda- mental valency, of each element to be the presence in varying proportions of a sub- element of atomic weight T }o or -0089. De- noting this by /x, then the univalent elements contain I/UL, the bivalent Z/UL, the tervalent 3/x, and so on. The greater part of the mass of the atom is conceived as due to the polymerisation of an entity consisting of the hydrogen atom less /m.. Denoting this by U, then hydrogen is H-f/x, lithium yH-f/x, carbon i2H-f-4,u, oxygen I6H+2/X, neon 20H. The^atom of hydrogen is thus made up of 9989 -[--0089= i -0078 : oxygen is similarly 9989 x i6 + -oo89 x 2 = 16-000. The author has calculated the whole of the atomic weights in this manner, and the agreement with the recognised figures de- rived from the results of experiment is quite * /. Chem. Soc. t xcv, 1752 (Nov., 1909). 1 06 THEORIES OF EVOLUTION remarkable, a difference not exceeding -04 being found in the great majority of cases. It is obvious that the resultant value for the calculated atomic weight will depend on the valency assumed for the element, and in the author's table certain elements seem to have been treated rather arbitrarily. Thus S has only 2/x assigned to it, P, As, and Sb only 3/x each, while Cr has 6yu, and V has 5^. On the whole, whatever may prove to be the physical significance of Moir's hypo- thesis, his numerical results are far more satisfactory than the earlier attempts to express by more or less complicated formulae the relationship between the atomic weights. The majority of such formulae contained constants or variables, which for the most part had no physical or chemical signifi- cance. Moreover, the atomic weights, ac- cepted as the result of refined modern ex- periment, are in some few cases substantially different from those of thirty to forty years ago. 107 CHAPTER V SPECULATIONS "!F we be curious to know what matter is, we plunge at once into that deep which surrounds us on every side, and which never yet was fathomed by human intellect. " With regard to its ultimate constitution we cannot hope to attain to a clearer conception than that which presented itself to the comprehensive but humble mind of Newton, and that transcendent philosopher has thus embodied the result of his patient investigations : "'It seems probable to me that God in the beginning formed matter in solid massy, hard, impenetrable, movable particles, of such sizes and figures and with such other properties, and in such proportion to space as most con- duced to the end for which He formed them ; and that those primitive particles being solids are incomparably harder than any porous bodies compounded of them ; even so very hard as never to wear or break in pieces, no ordinary power being able to divide what God Himself made one in the first creation.'" DANIELL'S Chemical Philosophy (1843), p. 7. the foregoing chapter it appears 1 that modern ideas as to the genesis of the elements, and hence of all matter, stand in strong contrast with those which chiefly prevailed among experimental philo- sophers from the time of Newton, and seem 1 08 SPECULATIONS to reflect in an altered form the speculative views of the ancients. Assuming the possibility of the evolution of matter, as we know it, from a primal essence, several questions require to be con- sidered in order that the process may be pictured in terms of those forms of energy and those forces with which we are familiar. The earliest stages are too difficult, and must be passed over without an attempt at ex- planation ; for supposing a protyle, it is im- possible to say what led to the first differ- entiation into discrete parts ; and if all were alike in mass and movement, what impressed one set of particles with the property of assuming the state called positive while another set acquired the power of becoming negative electrically. Nor can we say whether electricity is itself something super- added to matter or whether it is matter it- self. Facts now at our disposal show that all matter is resolvable into the two parts, positive and negative, and the elements of 109 THE ELEMENTS the chemist of which all terrestrial matter consists are capable of being brought into a common scheme. The questions which admit of discussion are concerned with the relative probabilities of the different possible views as to the order in which these elements have been evolved, and the manner in which the negative protyle may be supposed to have co-operated with the positive toward the formative process. The question may also be considered whether it is probable that the elements have all been formed one after another in an order corresponding with the order of their atomic weights, and whether the process should be supposed to be of a generally uniform character throughout, or whether it is not justifiable to imagine that this uniformitarian view should be modified so as to admit the occasional operation of energy derived from sources other than those immediately and continuously con- cerned in the formative process. In order to enter on this discussion a few no SPECULATIONS facts should be recalled. The study of the radio-active elements, so far as it has gone within the few years that these substances have been recognised, has revealed one fact which is of great importance in connec- tion with the question of evolution. Radium, the best known of these ele- ments, is a metal the compounds of which resemble those of barium, but with the im- portant difference that it undergoes spon- taneous disintegration and decay, giving off particles of helium and a gaseous emanation which is ultimately resolved into helium and nothing else. What may be the residue left after the escape of the emanation from radium is not known with certainty, but it is believed to be one of the common metals. Radium, then, seems to contain within its atom a store of helium, the escape of which is attended by the liberation of an enormous amount of energy, so that the process may be compared with the decomposition of an " endothermic " chemical compound, in THE ELEMENTS with the qualification that the energy liberated is vastly greater than the amount liberated in any known chemical process. The constitution of radium is probably imitated by that of the other radio-active elements, actinium, polonium, etc., which, in respect to their chemical characters, are at present very imperfectly known, though there is reason to believe that they agree in possessing metallic characters. It seems to be established that metals of the ordinary type, such as potassium and the rest of the alkali metals, and others under special conditions, emit /3 rays similar to those of radium, which are attributed to the ex- pulsion of negative particles or electrons ; but there is at present no evidence that any- thing corresponding to an emanation resolv- able into helium escapes from them. It is conceivable that the instability of the radio- active elements may be due not only to their larger atomic mass, but to their peculiar constitution, the helium atoms expelled I.I 2 SPECULATIONS from radium, for example, being apparently ready formed within its larger atom, which would thus possess a sort of grained struc- ture, while in ordinary inactive stable ele- ments the electrons may be supposed im- bedded in the positive shell in positions from which they are not so readily dislodged. From the experiments of J. J. Thomson it appears that helium itself, like other gases, under the influence of the electric discharge gives negative corpuscles in the rays from the cathode, and positively charged particles simultaneously. Hence helium and the rest of the argon group are also probably con- stituted of positive and negative protyles, and the chemical inactivity of these ele- ments must be attributed to the peculiar intimate kind of association in which the two protyles exist.* * It would seem, however, that dissociated as they are under the action of electricity some chemical activity ought to be induced under these circumstances, and the reason why it has not been observed is that the amount thus dissociated is too small to be recognisable by any ordinary chemical agent. Ramsay, " L'Helium," Ann. Chim. [7] xiii, 38 (1898). I II 3 THE ELEMENTS Supposing Crookes' figure of eight to be accepted as the basis of a diagrammatic representation of the changing conditions under which the protyle condenses into atoms of matter it seems to need some modification. For inasmuch as condensa- tion, mechanical or chemical, is always accompanied by loss of energy to the system, every time that an element is formed an evolution of energy takes place, which, whether it take the form of heat or of electrical potential, must temporarily arrest the process of condensation till it has in some way become dissipated. It would also seem probable that, supposing a similar constitution for both, the formation of a large atom from protyle must be attended by the separation of a larger amount of energy than is the formation of a small atom. The successive steps of the process by which the elements are severally pro- duced according to the order of their atomic weights must therefore become slower and 114 SPECULATIONS slower, so that the intervals occurring be- tween the deposition of one atom and the next will become necessarily greater. Hence the distribution of the elements at equal intervals along the curve does not seem to harmonise with probability. The periods represented by the lemniscate track should increase in descending so as to bring in the element of time. It might be argued that if the elements of high atomic weight were formed slowly they would probably be more stable than those of smaller atomic weight formed more quickly, or at least equal in stability. This, however, does not seem to be the case, for the radio-active elements are all found among those of the highest atomic weight, so that some elements seem to have been formed hastily with an internal constitution not framed to assure permanence. It is therefore conceivable that the process of condensation did not always result in the formation of one element at a time ; a non- S THE ELEMENTS metal may have attended the production of a metal, or an element of high atomic weight may have come into existence simultane- ously with one of low atomic weight. An analogy is afforded by the processes of polymerisation of ordinary matter. For example, aldehyde C 2 H 4 O is convertible simultaneously into paraldehyde (C 2 H 4 O) 3 and metaldehyde (C 2 H 4 O) n , the change being attended by evolution of heat and con- traction. The condensation of protyle must, how- ever, be imagined as an interaction of posi- tive and negative particles, and an analogy, as far as it goes, might be found in the action of hydrogen on sulphur at an elevated temperature. Here we should have balanced actions : H 2 + S ^ ^ H 2 S nS ^~^ S n allotropic sulphur ; while the proportions formed, and the state of equilibrium maintained, are under the operation of the law of mass action. 116 SPECULATIONS We have already seen that of the known elements those which exhibit the metallic character are by far more numerous than those which are definitely non-metallic, and this seems to suggest that probably the constitution of the metals is generally more stable than that of the non-metals. But inasmuch as radium and the other radio- active elements, which are all of high atomic weight, are actually in process of spontaneous disintegration, they must differ from the rest in regard to internal constitution, and a suggestion has already (p. 112) been made as to the possible cause of their instability.* These elements, radium, actinium, polonium, are now all that remain of what was prob- ably in earlier stages of creation a more * It is not very clear why the disintegration should take place atom by atom, or, in other words, what it is that determines the explosion of one atom rather than another in its immediate neighbourhood, and having presumably the same constitution. It does not appear to be the result of mass action, for the products of disintegration pass away as soon as liberated, and do not seem to accumulate in the residue. 117 THE ELEMENTS numerous family of similarly radio-active bodies. For as radium yields helium the first term of the chemically inactive series, it seems legitimate to suggest a similar origin for neon, argon, krypton, and xenon. It is improbable that any traces of these ele- ments remain among the constituents of the earth. Their formation must have been the result of a process of condensation of the hypothetical protyles, but they were probably formed at an early stage of the process of evolution, as by-products in the operations from which issued the common permanent elements of relatively low atomic weight, which at this day constitute the basis of all terrestrial and most celestial things. The disintegration of these large atoms must have been attended by the liberation of large stores of energy greater even than that which is known to be set free from radium> and this liberation of energy must have been a disturbing agency, possibly nS SPECULATIONS intermittent, which may well be credited with some of the irregularities in the develop- ment of the scheme of creation which- re- sulted in the elements of our system. Substances of this character are already known among the products of the disinte- gration of radium, all of which are rather short-lived, but probably have a definite atomic weight. Of these the most important is the " emanation " from radium. This is a chemically inert gas which may be con- densed at a low temperature to a solid, and the period of half transformation of which is 3-8 days. It is ultimately resolved wholly into helium, and in the process of trans- formation gives out heat at the rate of 75 gram calories per hour, at its maximum, for the emanation released from i gram of radium.* This enormous liberation of energy in so concentrated a form may well be credited with the power of arresting or promoting, according to circumstances, a * Rutherford's Radio-activity^ 2nd ed., p. 431. THE ELEMENTS process of condensation proceeding in any material in contact with the emanation. Such a change on a large scale may have affected the nature of the products hypo- thetically resulting from the condensation of " protyle." To some such agencies may be reasonably attributed Mendeleeff's eighth group, the anomalous atomic weight of tellurium, the peculiarities of the cerium group of earth metals, and even the eccentricities of in- dividual elements such as mercury or thallium. Turning now to the elements o{ lowest atomic weight, the spectroscope has shown that hydrogen and helium appear more widely distributed in nebulae and stars than elements of greater atomic weight, such as calcium, magnesium, and iron. But the evidence concerning the existence of other elements, to which names such as asterium, nebulium, and coronium have been given, is, to say the least, very uncertain. Hence 120 SPECULATIONS the hypothesis of Jessup (p. 84) is founded on a very insecure basis. The latest an- nouncement of the kind comes from Pro- fessor Wolf, of Heidelberg, and attention has been drawn to it in Nature by Professor Brauner, of Prague.* From experiments on centrifugal rotation of a mixture of gases, by Lobry de Bruyn and others, it appears to be established that the constituents of such a mixture are partially separated by very rapid rotation, the gas of higher density, and hence greatest molecular weight, present becoming concentrated in the periphery as the radius of rota- tion is increased. Wolf finds that the ring nebula in Lyra consists of four gases, which, owing to rapid rotation of the ring, have become separated into four different layers. The smallest ring A, representing the innermost part, is composed of an un- known gas. The next larger, B, consists of hydrogen ; the next, C, external to B, con- * Nature, Ixxx., 158 (April 8, 1909). 121 THE ELEMENTS sists of helium ; while the outermost ring, D, consists of another unknown gas. So that A must be lighter than hydrogen, and D must be heavier than helium. Professor Brauner reminds us that Men- deleeff supposed the existence of an element with atomic weight 0-4, or one-fifth the density of hydrogen, but belonging to the inactive series like helium ; and this ele- ment he provisionally identified with Young's coronium, a gas hitherto recognised only in the sun. The nature of the denser gas leaves room for speculation. Brauner himself suggests a gas with a smaller atomic weight than helium, but having a larger molecular weight. This implies that its molecule must be composed of more than one atom, and it therefore would not fall into the argon family. I have recently * suggested the possibility of an element standing toward the halogens in the same * "Mendele'eff Memorial Lecture." /. Chem. Soc., xcv, 149(1909). 122 SPECULATIONS relation, as regards atomic weight, as hydrogen to the alkali metals. This would have a density about 2-8, and a molecular weight 5-4. Mendeleeff, in his latest specu- lations (1905) concerning the possibility of still undiscovered elements, suggested the existence of an element of the halogen group with atomic weight about 3. The non-metals are notoriously more difficult of recognition than elements of metallic or metalloidal character. But the identifica- tion of oxygen and nitrogen, and recently of sulphur,* in some of the hotter stars lends some support to the idea that the spectrum observed by Wolf may be due to an element of this kind. In Mendeleeff's latest table of the ele- ments (pp. 46-47) y is supposed to be a very light gas of the same monatomic con- stitution and inactive character as helium, and may be identified hereafter with " coron- ium," which is found in the sun's coronal * Lockyer, Froc. R. Soc ty Ixxx, 50, 1907. 123 THE ELEMENTS atmosphere. This gas, according to Men- deleeff, would have a density about 0-2, and therefore a molecular weight and atomic weight about 0-4, or about one-tenth that of helium. The x in this table is the " ether " of the physicist for which Mendeleeff, disregarding conventional ideas, assumed a molecular structure with an extremely small density and atomic weight. The properties attri- buted to the ether, which pervades all space and penetrates through all ordinary matter, show, however, that, whatever it may be, it is not a gas. x may stand for the present purpose to represent the hypothetical posi- tive protyle, while x may be placed in the corresponding "position at the end of Series as the symbol of the other, negative, parent of common matter. Having now cleared the ground, we may proceed to consider what are the most probable lines along which evolution may be supposed to have been accomplished, 124 SPECULATIONS and a few general principles must be laid down before attempting to display any general scheme. i. Mendeleeffs periodic law is applicable to a large number of elements, though, strictly speaking, the scheme does not pro- vide for all. The properties of the elements are undoubtedly determined partly by atomic weight, but they are also influenced, perhaps in a greater degree, by what must be supposed to be their internal constitution, or the arrangement of the electrons within the atom. It does not, however, follow that they were evolved one after another in the order shown in the table. The curves of Reynolds and Crookes are much too simple in assuming the operation of uniform physical conditions from first to last. Whereas from the irregularities and anom- alies noticeable in so many parts of the scheme it seems more likely that the pro- cesses of construction assumed several forms, and were liable to interruption or accelera- 125 THE ELEMENTS tion from causes which have been already referred to. Thus while a large number of the elements may have been produced by simple con- densation together of positive and negative protyle, others were probably formed by disruption of big atoms, a kind of depoly- merisation, or again by special and peculiar processes due to the disturbing effects of energy introduced by such disruption. 2. The radio-active elements were prob- ably formed by direct condensation of protyle, but this condensation occurring without the loss of so much energy as may be supposed to occur during the formation of the common stable elements, an unstable endothermic structure resulted. 3. Condensation probably occurred in the order of the natural families, that is, down the vertical columns of Mendele"efFs groups, so far as the cases of closely allied elements are concerned. It is, for example, highly probable that sodium was formed immedi- 126 SPECULATIONS ately after lithium, followed by potassium, rubidium, and caesium. The idea that the metals owe their obvious community of characteristics to the presence in them of a common constituent is of no recent date. It was the radical principle of the ancient doctrine concerning transmuta- tion. The alchemists believed that the " base " metals and the " noble " metals differed only by reason of the presence in the former of some kind of disease or im- purity which obscured the pure metallic qualities of their noble relatives silver and gold. The notion was quite seriously referred to by Davy in the early part of the nine- teenth century, and is one which has been repeatedly revived. The families of negative elements were probably produced in a similar manner, though in this case the modification of character which attends increase of atomic weight is more marked, the change amount- ing to a partial change of function, as may 127 THE ELEMENTS be seen by comparing together sulphur and tellurium, or still more obviously phosphorus with bismuth, or silicon with lead. 4. It seems to be agreed that a limit is set to the magnitude of atomic weight in consequence of the general instability of elements with large atoms. Uranium stands at present at the end of the known series. It is believed to be undergoing spontaneous disintegration, but it is evidently more stable than radium, which is supposed to be derived from it, and this affords direct evidence that constitution has more to do than mere mass in determining the character of an atom. Here it may be remarked that both radium and polonium seem to be in process of forma- tion now and always from uranium, and in the absence of information as to the nature of the change a question arises as to the real nature of the primal stuff. Is it uranium itself or something of still higher atomic weight existing in small quantity in uranium? 128 SPECULATIONS 5. The elements of the argon group were pretty certainly products of devolution from substances of high atomic weight, most of which have become extinct so far as this earth is concerned. Probably a few other elements of the ordinary type, e.g. lead, represent the residues of this disruptive operation. 6. Evolution of some families may be assumed not to have proceeded by the simple accretion of matter or electrons as suggested (3), but starting from one or other of the typical elements the course of condensation may have followed more than one path, while preserving the same general character of product. Take the case of the iron metals, for example. The ordinary arrangement of the metals in the periodic scheme separates iron, nickel, and cobalt, on the one hand, from copper and zinc ; and on the other from manganese and chromium. But these metals form an isomorphous group in which the isomorphism is repeated in the K 129 THE ELEMENTS salts corresponding to the ferrous, as well as to the ferric salts and to the chromates as far as they go. The valency of manganese has been mis- represented on the basis only of the isomor- phism of permanganate with per chlorate in order to force it into the halogen group. Placing these metals in the following order the reader is reminded of their close inter- relationship, as well as the gradual decline of valency, and of the permanence of the higher oxides among them. Valency. At. Wt. Density. At. Vol. Chromium . II III IV VI 52-0 6-9 7 *S Manganese . II III IV VI 54-9 6-85108-0 7 '4 Iron , ; , II III IV VI 55'8 7*8 7 i Cobalt II III IV VI ? 59 -o 87 6 8 Nickel II III 587 8-8 6 7 Copper II 63-6 8-9 7 i Zinc . . II 65-4 6-9 9-5 Cobalt appears to be related to manganese much in the same way as nickel to iron. The only metal from which the whole group could be descended is aluminium, which is isomorphous with iron in the ferric 130 SPECULATIONS state, but has the lower density, 2-6, and lower melting-point, 650, which agree with its lower atomic weight, 27-1. In the following table, which preserves the chief features of the periodic scheme, an attempt is made to indicate the probable order of evolution. The two lateral divisions marked A contain the families which are supposed to have been formed by direct process of condensation, the successive ele- ments being produced one after the other, as indicated by the arrows. The middle division, B, contains elements which for the most part are supposed to have been pro- duced indirectly or irregularly. Concerning the " rare earths," no attempt has been made to trace their origin, though it is pos- sible that they are descended from alumin- ium through scandium, yttrium, and lan- thanum, x stands for the hypothetical pre- fluorine referred to on pages 122-3. The diagram speaks for itself, but it may be as well to point out that the first eight THE ELEMENTS elements, including hydrogen, are supposed to be formed by direct union of x and ~%, the two hypothetical protyles ; and this is in- dicated by the lines terminating in each case in an arrow-head. The next seven are sup- posed to result from condensation upon the body of the already formed atom of more protyle, positive or negative, according to the structure and requirements of the atom which forms the nucleus. According to this view sodium was formed from lithium, magnesium from beryllium, and so forth. A similar development is sup- posed to proceed downwards into the suc- cessive families ranged in the departments of the table marked A, the order of succes- sion being indicated by arrow-heads. Where no arrow-head is shown no opinion is ex- pressed, although the line may pass through a symbol. For example, tantalum is sup- posed to be the lineal descendant and de- rivative of columbium (niobium) without im- plying any view in regard to praseodymium. 132 > is.) (i 5 Extinct ? \ _ Em? IX Ae 9 V w Co ,Cu >^ :, ^ * ^ *^ ^u___^ 1 :a n : n (Ga) (Tk) (Dy) ( ? -r) C ) _ Ir >R ^ | ^S 1 v Extinct? Exlmct? hwl ? R,? Te face page 132 SPECULATIONS In the division B the elements are repre- sented as being formed in the order repre- sented for the most part along horizontal lines, in most cases the development pro- ceeding in the direction of increasing atomic weight. The most doubtful case is the deri- vation of ruthenium from iron, but the series Ru - Ro - Pd - Ag is parallel to the series Os - Ir - Pt - Au How the elements on the last horizontal line came into existence it is not possible to suggest in a diagram, but the probable process has been sketched in par. 2. From what has gone before it seems prob- able that the chemical elements, and hence all material substances of which the earth, the sea, the air, and the host of heavenly bodies are all composed, resulted from a change, corresponding to condensation, in something of which we have no direct and intimate knowledge. Some have imagined this primal essence of all things to be iden- THE ELEMENTS tical with the ether of space. As yet we know nothing with certainty, but it is thought that by means of the spectroscope some stages of the operation may be seen in progress in the nebulae and stars. This affords a wide field for speculation, and possibly our knowledge may never become more certain than it is to-day. In some departments of thought, however, we must be content with circumstantial evidence, which, if only sufficient in amount, may serve as a satisfaction to the mind. Until quite recently the elements of the inorganic world were supposed to be fixed, immutable with the lapse of ages or under the mighty forces concerned in the making of worlds. But within a few years we have learned that some atoms, at any rate, are not permanent, but are continually crum- bling away. There seems in all this no obvious parallel between the changes which have led to the formation of the elements and the evolution SPECULATIONS of living beings. In organic evolution spon- taneous variation of form and function, and the struggle for existence, leads to the survi- val of the fittest. If some of the elements are actually exposed in nature to the attack of a or ft particles, atoms of helium, or some- thing else, moving with immense velocity comparable with that of light, they may be breaking down. As yet it is impossible to say whether all may not be suffering a slow waste which in the long run must lead back to the primal chaos. " How real existence is to be studied or discovered is, I suspect, beyond you and me. But we may admit so much that the know- ledge of things is not to be derived from names. No ; they must be studied and in- vestigated in themselves " (Socrates, in Plato's Cratylus; Jowett's translation). INDEX Abegg, 103 Alkali Metals, 60 Argon Group, Origin of, 118, 129 Asterium, 83 Atomic Theory, 8 Volume, 43 Weights Standardised, 17 Interrelations among, 25> 5 6 Avogadro's Hypothesis, 9, 17, 19 Berzelius, 92 Boisbaudran, L. de, 49 Brauner, 121 Bruyn, de, 12 1 Cannizzaro, 17 Carnelly, 93 Chancourtois, B. de, 34 Coronium, 122, 123 Corpuscles in rings (table), 100 Corpuscular Theory of Matter, 95 Crookes, W., 72, 73, 79, 8r, 125 Crookes' Figure of Eight, 74, H4 Dalton's Atomic Theory, 8 Davy, 127 Definite Proportions Law, 7 Divisibility of Matter, 5 Doebereiner, 27 Dulong and Petit, Law of, 20 Dumas, 31 Element, Definitions of, 3 Elements arranged by Od- ling, 32 de Chancourtois, 34 Newlands, 36 Mendeleeff, 40, 46 in order of Atomic Weight, 52 Natural Families of, 57 Emanation from Radium, 119 Energy absorbed in de- composition, 5 137 INDEX Energy evolved in combina- tion, 5 Equivalents determined, 1 8 Ether (Mendeleeff), 124 Evolution, Theories of, 71 Gay-Lussac's Law of Vol- umes, 9 Genesis of the Elements (Crookes), 72 Gerhardt, 17 Gladstone, 38 Graham, 78 Halogens, Properties of, 61 Helium, chemical activity, "3 Helium in stars, 83 Hoflf, Van't, 15 Homologous series, 27 Hydrogen in stars, 83 Iron Group, 130 Isomorphism, 21 Jessup, 82 (Table of Elements), 86 Kalidium, 94 Kelvin, 8 1 Law of Octaves, 36 LeBel.J. A., 14 Lockyer,J. N., 83 Mendeleeff, 39, 45, 48, 51, 66, 71, 88, 124 Metals, characters of, 90 Base and Noble, 127 Meyer, L., 42 Meyer's Curve, 44 Moir, J., 106 Nebula in Lyra, 12 1 Newton, 108 Non-Metals, characters of, 90 Octaves, Law of, 36 Odling, 17, 32 Oxidium, 94 Palmer, 93 Pasteur, 14 Periodic Law, 51 Periodic Scheme, 22, 66, 70, 125 Polymerisation, 116 Prefluorine, 94, 122, 131 Proto-beryllium, 85 Proto-boron, 85 Protyle, 72 Prout's Hypothesis, 25 Radio-active Elements, for- mation of, 115, 126 Radium, in, 117 Ramsay, 105 Rare Earths, 65 I 3 8 INDEX Reynolds, E., 73, 125 Speculation, 108 Spiral representing Evolu- tion (Crookes), 74 Stars, Components of, 83 Stereochemistry, n Stoney, Johnstone, 104 Tartaric Acids, 13 Thomson, J. J., 81, 94, 113 Uranium the limit, 56, 128 Valency, 59 Van't Hoff, 15 Vis tellurique, 35 Williamson, 17 Wolf, 121 Young, 122 WILLIAM BRENDON AND SON, LTD. PRINTERS, PLYMOUTH " Presenting suggestive ' living' thoughts on subjects of vital interest." :: :: :: The Times. Harper's Library of Living Thought Prof. ERNEST A. GARDNER RELIGION AND ART IN ANCIENT GREECE Just Issued This volume enables the modern reader to realise the relation between religion and art under the most favourable conditions the con- ditions which existed in ancient Greece and it enables him to understand what representations of the gods actually meant from a religious point of view, both to the common people and to the leaders of thought. The subject is treated historically, not only in the tracing of artistic development, but in the study of the different stages of idolatry, ideal conceptions, individualisation, and symbolism, and their effect upon the religious feeling of various periods. It shows the working of interacting influences in the development of characterisation in sculpture and in the anthropomorphism of the religion of ancient Greece. " Not reprints of the classics, but the work of living writers." :: :: :: Evening Standard. " ' Thought,' in the most exalted sense of the word, and certainly alive." :: :: Nature. Harper's Library of Living Thought C. H. Hawes f M.A., and Harriett Boyd Hawes, M.A. CRETE, THE FORERUNNER OF GREECE " The wondrous story of a great civilisation which flourished before Abraham was born, and left behind a memory of itself in the Arts of Ancient Greece and in the traditions of a golden age and a ' Lost Atlantis.'" Evening Standard. Prof. Paul Vinogradoff ROMAN LAW IN MEDIEVAL EUROPE " Really an important and suggestive study which no student of history can afford to neglect." Manchester Guardian. " This brilliant little book is worthy of a place beside Maine's 'Ancient Law,' to which it forms an invaluable sequel, the one telling the wonderful story of evolution, the other a still more wonderful story of decay and reintegration. " Glasgow Herald. Algernon Charles Swinburne THREE PLAYS OF SHAKESPEARE The last work of the author published before his death. "These are noble essays, in a style which is as youthful, vigorous and uncompromising as in the days 'Before Sunrise.'" Westminster Gazette. "The best thinkers of the day on matters of vital importance and profound interest." Liverpool Post. Harper's Library of Living Thought Now Ready Algernon Charles Swinburne THREE PLAYS OF SHAKESPEARE Leo Tolstoy THE TEACHING OF JESUS W. M. Flinders Petrie PERSONAL RELIGION IN EGYPT BEFORE CHRISTIANITY Sir Oliver Lodge, F.R.S. THE ETHER OF SPACE. Illustrated Prof. William Wrede (University of Breslau) THE ORIGIN OF THE NEW TESTAMENT Prof. C. H. Becker (Colonial Institute, Hamburg) CHRISTIANITY AND ISLAM Prof. Svante Arrhenius (Nobel Institute. Stockholm) THE LIFE OF THE UNIVERSE. 2 vols. Illustrated Prof. Arnold Meyer (University of Zurich) JESUS OR PAUL? Prof. D. A. Bertholet (University of Basle) THE TRANSMIGRATION OF SOULS Prof. Reinhold Sezberg (University of Berlin) REVELATION AND INSPIRATION Prof. Johannes Weiss (University of Heidelberg) PAUL AND JESUS Prof. Rudolf Eucken (University of Jena) CHRISTIANITY AND THE NEW IDEALISM Prvf. P. Vinogradoff (Oxford University) ROMAN LAW IN MEDIAEVAL EUROPE Sir William Crookes, LL.D., F.R.S. DIAMONDS. Illustrated C. H. Hawes, M.A., and Harriet Boyd Hawes, M.A. CRETE, THE FORERUNNER OF GREECE Sir William A. Tilden, F.R.S. THE ELEMENTS : SPECULATIONS AS TO THEIR NATURE AND ORIGIN Prof. Ernest A. Gardner (University of London) RELIGION AND ART IN ANCIENT GREECE Other volumes will be duly announced Please write for a prospectus HARPER & 45 Albemarle Street, London, W. May, igio. BROTHERS : : Franklin Square, New York. UNIVEESITY OF CALIFOENIA LIBEAEY, BEEKELEY THIS BOOK IS DUE ON THE LAST DATE STAMPED BELOW Books not returned on time are subject to a fine of 50c per volume after the third day overdue, increasing to fl.OO per volume- after the sixth day. Books not in demand may be renewed if 'application is made befo.re expiration of loan period. ftlAR 11 1932 JAM 29 19 30 No '53 DX 353 HI RECD 55TT 75m-7,'30 YA 02364 T Tr 1 o 21666G