HISTORICAL INTRODUCTION.1 EVER since the establishment of the atomic theory by Dalton and Berzelius it was felt among chemists that there must be some relation between the atomic weights of the different elements and their properties. It was recognized very early that there exist groups of elements possessing related chemical and physical properties, and one of the earliest attempts to bring out this point is due to Dobereiner. In 1829 he tried to show that “many elements may be arranged in groups ()f three, in each of which the middle element has an atomic weight equal or approximately equal to the mean of the atomic weights of the two extremes.” As illustrations of this method of arrangement may be mentioned the following groups: Li, Na, K; Ca, Sr, Ba; and Cl, Br, I. Passing over briefly the memoirs of Cooke and Be- guyer de Chancourtois, we come to the “law of octaves” nunciated by J. A. R. Newlands in 1864. He drew Group I. Group II. Group III. Group IV. Group V. Group VI. Croup VII. Group VIII. &EMI&. BH« SH' SH' RH S'O SO R'O' SO1 S'O' SO' R'O' SO' 1 H-l X Li - 7 Be -9'4 B - ll C - 12 N-14 0-16 F - 19 3 Na - 23 Mg - 24 AI -:/7'3 Si = P - 31 S - 32 CI - 3S'S 4 K - 39 Ca - 40 — - 44 Ti - 48 V - 51 Cr - 52 loin - 55 Fe - 56 Co - 59 Ni - 59 Cu - 63. Ii (Cu - 63) Zn - 65 68 72 Aa - 75 So - 78 Br - SO II Rb - 85 8r - 87 IYt - 88 Zr - 90 Nb - 94 Mo - 96 — - 100 Ru - 104 Rh - 104 Pd - 106 As - 108 7 (As - 108) Cd - 112 In - 113 Sn - 118 Sb - 122 Te = 1:151 1 - 127 _ _ _ _ Co - 133 Ba - 137 ?Di - 138 We - 140 — — — II (-) 10 — — IEr - 178 ?La - 180 Ta = 18:1 W - 184 - Os - 105 Ir - 197 Pt - 198 Au -.199 11 (Au - 1119) HS - ZO T1 -:104 Pb -:107 Bl - — — 12 — — — Th - U- 240 — Fig. 1. Periodic Table as Arranged by Mendelejeff attention to the fact that “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,” and thus made_the most distinct advance toward a system of classification of the elements that had yet been accomplished. It is, however, to the Russian chemist, Mendelejeff, that chemistry owes the system of classification of the elements which is based on the recognition of this fundamental fact: “that the properties of the elements and the properties and compositions of compounds vary periodically with the atomic weights of the elements.” This principle, known as the Periodic Law, was enunciated by Mendelejeff in two memoirs published in 1869 and 1871, respectively, and the arrangement of the elements, based on this law, which was finally adopted by him is illustrated in Fig. 1. While a discussion of this law may be found in almost any text-book on chemistry, a few remarks of a general nature may not be out of place in this connection. Mendelejeff arranges the elements into series and groups. In each series the order of the elements corresponds to increasing atomic weights, and accompanying this change in atomic weight there is evident a gradual variation in all the properties of both the elements and their compounds. On the.other hand, the arrangement in groups exhibits the periodical recwrrence of elements possessing fairly analogous properties. The change in valency, as exhibited by the formulro of the oxides and hydrides, is probably one of the most striking facts brought out by the periodic arrangement of the elements. From the univalent elements like H, Li, Na, etc., the valency for oxygen increases regularly until in compounds like OsO, the elements exert a valency of eight. The maximum valency for hydrogen appears to be four, and while the valency for oxygen increases from Group I to Group VIII, that for hydrogen decreases in the same manner from Group IV to Group VIII. The compounds exhibit a gradation in properties quite similar to that exhibited by the elements themselves. Thus Na, O is strongly basic, MgO less so, Al, O3 combines with acids to form salts and with alkali hydrates to form aluminates, that is, it ads as an anhydride of both acids and bases. In Si02 we have a weak acid anhydride, while the acids formed from P20s, SO and Cl, O, range in strength in the same order. ATOMIC VOLUME AS A PERIODIC FUNCTION OF ATOMIC WEIGHT. Probably the best illustration of the significance of Mendelejeff's Periodic Law can be conveyed by plotting some property of the different elements against the atomic weight. In Fig. 2, which is taken from Holle- man's Inorganic Chemistry, the atomic volume (specific gravity divided by atomic weight) has been plotted as ordinate with the atomic weights as abscissro. It will be observed that elements possessing similar chemical and physical properties occupy similar positions on the curve. In mathematics a periodic function is one which returns to the same value for definite increments of the independent variable. From Fig. 2 it is evident that we can in a similar manner state that the atomic volume is a periodic function of the atomic weight. The specific heats of the elements when plotted as ordi- nates against the atomic weight show a similar periodicity of maxima and minima, and the same can be stated for other properties. APPLICATION OF PERIODIC LAW TO DETERMINE ATOMIC WEIGHTS. One of the most important applications of the Periodic Law suggested by Mendelejeff was the determina tion of atomic weights from the properties of the elements. In other words, he stated as a fundamental axiom that the atomic weight of element must determine its properties. He illustrated this conclusion by prophesying in detail the properties of three unknown elements which he named eka-boron, eka-alu- minium, and eka-silicon, and to which he assigned the approximate atomic weights 44, 68, and 72, respectively. His predictions were subsequently completely verified by the discovery of the elements scandium (eka-boron), gallium (eka-aluminium) and germanium (eka-silicon). It must be observed that without the assistance of the Periodic Law the exact determination of the atomic weight of an element, whose compounds are all nonvolatile, becomes a matter of extreme difficulty. Thus a chemical analysis of the oxide of indium shows that the element has the equivalent weight 38, that is 38 parts by weight of indium are equivalent to 1 part by weight of hydrogen. At the time when Mendelejeff published his papers the atomic weight of this element was taken to be 76 and the formula of the oxide was assumed to be InO. A study of the properties of this oxide and of the metal itself, from the standpoint of the Periodic I, aw, led Mendelejeff to assign it to Group III, along with B and Al. Consequently the oxide must have the formula InO3 and the atomic weight must be about 114. discrepances in the periodic table. It was already observed by Mendelejeff that a discrepancy exists in the case of tellurium and iodine. According to order of atomic weights iodine should come before tellurium; but even the most superficial investigation of the properties of these elements and of their compounds shows that iodine belongs to the chlorine family, while tellurium closely resembles sulphur and selenium. Mendelejeff therefore argued that the atomic weight of tellurium ought to be smaller; but in spite of the most careful and most elaborate investigations undertaken in this direction, the results have always led to the same conclusion. Similar discrepancies have been observed in the case of cobalt and nickel, and argon and potassium (see. “Rare Earths,” page 620). It will be shown in a subsequent section that these discrepancies disappear in the light of the most recent speculations. rare gases in relation to the periodic table. When the existence of the rare gases' was discovered an interesting question arose as to their place in the Periodic Table. As is well known, these gases were found to be absolutely inert chemically, thus differing radically from every other element known up to that time. Consequently they could not be placed in any of the known groups. However, by arranging them in a group to the left of Group I (see Fig. 4) they are shown as a natural transition from the elements of Group VIII to those of Group I. rare earths in relation to the periodic table. The group of elements known as the “rare earths” has presented an exceedingly interesting problem as regards their arrangement in Mendelejeff's system of classification. The elements of this group and their compounds resemble each other very closely in chemical properties; in fact, it is possible to separate them only because of slight differences in physical properties, such as solu- Fig. 2.—A graphical representation of the periodic variation of the actomic volumes of the elements with their atomic weight. © 1916 SCIENTIFIC AMERICAN, INC bility, melting point, or color; so that the process of isolating a salt of any one of the members of the group is a most laborious process, involving probably several thousand recrystallizations. Up to the present the existence of the following elements has been definitely determined: Atomic weight. Scandium Group: Scandium 44.1 Yttrium 88.7 Cerite Earths: Lanthanum 139.0 Cerium 140.25 Prreseodymium 140.6 Neodymium 144.3 Samarium 150.4 Europium.. 152.0 Ytterbium Earths: Gadolinium 157.3 Terbium 159.2 Dysprosium 162.5 Erbium 167.4 Thulium 168.5 ytterbium 172.0 Lutecium 174.0 With respect to the first four of the above elements, there has been no doubt as to what place they ought to occupy in the Periodic Table. When scandium was first isolated in 1879 it was recognized immediately as the element eka-boron whose properties had been prophesied by Mendelej eff. The position of yttrium and lanthanum in Group III as analogous elements to aluminium and scandium has also not been questioned. As cerium forms an oxide CeO. similar to SnO. and its salts resemble those of tin and germanium, it seems equally well established that this element belongs to Group IV. But up to the present time it has remained quite an open question as to the manner in which the other twelve elements should be arranged. Prof. Meyer has suggested that they should be grouped together in Gr(tup III between lanthanum and cerium, thus emphasizing the resemblance in chemical properties of the different elements constituting this group. This would, however, place lutetium, with an atomic weight of 174, before cerium whose atomic weight is 140. In view of the more recent work of Moseley on the high-frequency spectra of the elements, of which further mention will be made, the writer has tentatively arranged the rare earths as indicated in Fig. 4. They are thus made to come in below lanthanum and cerium and before tantalum. RADIOACTIVE ELEMENTS. The discovery of the radio-active elements has naturally led to the question as to what relationship they bear to the other elements in the Periodic Table. There could be no doubt about the position of elements like radium,-”thorium, and uranium which could be obtained in large enough quantities to determine their atomic weights and chemical properties, but up to the past year there was a great deal of speculation about the manner in which the other radioactive elements should be arranged, and it was only after an immense amount of careful investigation and ingenious deduction on the part of brilliant physical chemists like Soddy and Fajans that the whole situation was cleared up, and another epoch-making ehapter added to the history of the Periodic Law. It is largely with the conclusion reaehed by these investigators that the present paper is specially eoncerned. As is well known, the radioactive elements are characterized by a greater or less instability. After a certain average period of existence, which may range from over a thousand million years, as in the ease of uranium (U), to a millionth of a seeond, as in the case of RaGu the atom disintegrates spontaneously and yields an atom which possesses totally distinct properties. The disintegration is detected by the expulsion either of alpha' or of beta4 particles. Accompanying the expulsion of beta particles there is also observed in a number of cases, an emission of gamma rays. These are electromagnetic pulses of extremely short wave length (about 10-' centimeters) and are probably due to the bombardment of the atoms of the radioactive substance itself by the beta particles. As a result of the large amount of careful work which has been carried out during the past few years in investigating the relationship between the different radioactive elements and their transformation products, it has been concluded that there exist three well defined disintegration series whose starting points are uranium, thorium, and actinium, respectively. Fig. 3 illustrates diagrammatically the manner in which the members of these series appear to be related. When mesothorium II disintegrates it yields radio- thorium and as a beta particle is expelled during the transformation there is no change in atomic weight. Radiothorium is chemically allied to thorium and non- separable from it. These facts lead to the conclusion that radiothorium belongs to Group IV and mesothorium II must therefore belong to Group III. Passing to thorium X, we here again come to an element which is chemically similar to radium, thus placing it in Group II. The atom of thorium X expels an alpha particle and yields thorium emanation, a gas which is inert chemically, and condenses at low pressures between — 120 deg. Cent. and —150 deg. Cent. The emanation resembles therefore the rare gases of the argon group. Thorium emanation is the first member of the group of transformation products that constitute the thorium “active deposit.” They are indicated in Fig. 3 as thorium A, B, 0” 0, and D. The diagrams illustrating the aetinium and uranium series are self-explanatory. In a general way the three series are quite similar. The most noteworthy feature about these radioactive elements is the fact that individual members of each series appear to be chemically indistinguishable from certain members of the other series. Thus thorium B and radium B possess identical chemical properties. If it were not for the difference in period of existence of both substances it would be impossible to differentiate them. ISOTOPES. Soddy first drew attention to this and similar cases of radioactive elements that are chemically identical, and since they must occupy the same place in the Periodic Table he has designated them isotopes. Thus the elements uranium X” ionium, and radio-actinium are istopic. A similar example is furnished by the three emanations, and by radium and thorium X. A remarkable feature about these isotopes is that although they are chemically the same, they differ in atomic weights. In other words, we have here cases of elements that are absolutely inseparable by all chemical methods so far devised, and yet differ in ' that respect which has hitherto been taken to be the most important characteristic of an element—Its atomic weight. SODDY'S LAW OF SEQUENCE OF CHANGES. A comprehensive survey of the chemical properties of the di1rerent radioactive elements has led Soddy and Fajans independently to an interesting and extremely important generalization which enables them to assign these isotopes to their places in the Periodic Table. It will be remembered that an alpha particle is a helium atom with two positive charges. By its expulsion, therefore, the atom must lose two positive' charges, and the atomic weight must decrease by four units. Similarly, the expulsion of a beta particle means the loss of a negative charge or, what is equivalent, the gain of one positive charge; and since the mass of the beta particle is extremely small compared with that of the atom, there is practically no decrease in atomic weight. Now in the Periodic Table the valency for oxygen, an electro-negative element, increases regularly as we pass from Group 0 to Group VIII, while that for hydrogen, an electro-positive element, decreases, i. e., the electro-positive characteristic increases by one unit for each change in the group number as we pass in any series from left to right. Furthermore, in each group the electro-positive character increases regularly with increasing atomic weight. These considerations led Soddy and Fajans to this conclusion: The expulsion of an alpha particle from any radioactive element leads to an element which is two places lower vn the Periodic Table (and has an atomic weight Which is four units less) while the emission of a beta particle leads to an element which is one place higher up, but has the same atomic weight. It is possible, therefore, to have elements of the same atomic weight, but possessing distinctly different chemical properties, and, on the other hand, since the effect of the emission of one alpha particle may be neutralized by the subsequent emission of two beta particles, it is possible to have two elements which differ in atomic weight by four units (or some multiple of four) and yet exhibit chemically similar properties. As an illustration, let us consider the Uranium Series. Uranium I belongs to Group VI. By the expulsion of an alpha particle we obtain uranium !” an element of Group IV. This atom in turn disintegrates with the expulsion of a beta particle. Consequently uranium X2 must. belong to Group V. In this manner we can follow the individual changes that lead to the different members of the series, and by means of the generalization of Soddy and Fajans we cannot only assign to each element its place in the Periodic Table, but also its atomic weight, as has been done in Fig. 3. This generalization has been of material assistance in elucidating some of the difficult problems in the study of the disintegration series. More than this, it has led to the intensely interesting conclusion that the end product of each of the three radio-active series is an isotope of lead. The results of the most recent work on the atomic weight of lead are in splendid accord with this deduction, as it has been found that lead, which is of radio-active origin, has a slightly lower atomic weight than ordinary lead.' In a couple of cases the isotope has not been definitely isolated, but there can hardly be any doubt of its existence. Thus, the disintegration product of radium C2 must be an element of Group IV, but the evidence for its existence is very meager. NUCLEAR THEORY OF STRUCTURE OF THE ATOM. All these conclusions are in accord with an interesting theory of atomic structure that was first put forward by Rutherford and elaborated by Bohr, Moseley and Darwin. As this theory has been discussed at great length in connection with another series of articles8 we shall limit ourselves here to a few remarks on its essential points. Stated briefly, this theory assumes the atom to consist of a positively charged nucleus surrounded by a system of electrons which are kept together by attractive forces from the nucleus. “This nucleus is assumed to be the seat of the essential part of the mass of the atom, and to have linear dimensions exeeedingly small compared with the linear dimensions of the whole atom.” According to Bohr, the experimental evidence supports the hypothesis that the nuclear charge of any element corresponds to the position of that element in the series of increasing atomic weights. The chemical properties of the atom depend upon the magnitude of this nuclear charge; since, however, any given number of electrons may assume different configurations it is possible for two or more elements to exist having the same nuclear charge, but possessing different atomic weights. In other words, the possible existence of isotopes is deduced from Rutherford and Bohr's assumptions. The atomic weight thus assumes the role of a secondary characteristic; the important property of any element is its nuclear charge, so that by arranging the elements in order of increasing nuclear charge we ought to obtain a much better approximation to a periodic © 1916 SCIENTIFIC AMERICAN, INC 46 SCIENTIFIC AMERICAN SUPPLEMENTS 2089 January 15, 1916 MENDELEJEFF'S PERIODIC SYSTEM OF THE ELEMENTS Containing Atomic Weights, Atomic Nurhbers and Isotopic Radioactive- Elements Group 0 Group 1 EitO Group 2 EO Group 3 E2O3 Group 4 EOa EH4 Group'S EzOa EHa Group 6 EOa EH2 Group 7 E20y EH Group 8 EO4 He S.99 (S) H: Li 6.” (S)' Be* 9.1 (4) 11.00 (5). 12.00 N mSfU1 o 16.00 (8) 19.0 (9) Nei Ar 33.88 (18) (10) Na 23.00 (11) 39. 1O (19) Mg 24.” (US) A1 21.1 (18) Si 28.1 (14) 31.04 (15) 32.07 (16) Cl 36.46 (17) Ca 40.07 (eo). Sc 44.1 (11)' Ti IB.l (22). 61.0 («S) Cr 62.0. (I.) Mn .”93 «« Fe Co Ni 66.84 118.97 88.68 («?) (ss). Kr 82.92 (38) Cu 63.67 (S9) Zn, 68.37 (SO,) Ga 69.9 (81) Ge 72.8 (82) As 74.96 (88) Se 79.2 (S4) Br 79.92 (35) Rb 86.4 8 (37) Sr 87.63 (88) Yt 69.0 (ss) Zr 90:6 (40) Cbt 93.6 (41) Mo 96.0 (48) Ru Rh Pd 101.7 lQt.9 106.7 (44) ' (45) (48) Xe 180.2 (04) Ag 107.88 (47) Cd 112.40 (48) In 114.8 (49) Sn 119.0 (60) Sb 120.1 (61) Te' 127.8 (5«) 126.92 (6S) Cs 132.81 (55) Ba 137.37 (58) La 139.0 (67) Ce 14O.2S (68) Fig..—Arranged by the Research Laboratory of the General Electric Company. (4» arrangement of the elements. It so happens that In most cases the order, of increasing atomic weight coincides with that of increasing atomic number (nuclear charge), but tthis need not be. so in all cases. - HIGH RAEQUENCY SPECTRA' OF THE ELEMENTS. Bohr showed that there must exist a 'definite relation between the' charge on the nucleus and the frequency of the characteristic X-rays emitted by the substance. Moseley, therefore, has' measured the wave-lengths of the characteristic X-rays emitted by the different elements when these ' were 'made anti-cathodes in an X-ray tube and has determined, in' this manner, the atomic numbers of all]! the elements frbm aluminium, 13, 00 gold, 79. There appear to' be only' three elements in this range which have not been discovered by the chemist. ' “PERIODIC TABLE IN PRESENT FORM. The revised' form of Mendelejeff's ' Periodic Table which has been drawn up in Fig. 4 presents an attempt to embody the most recent results of the different lines of' investigation that have been discussed herein. Under each element is given the atomic weight' and the atomic number (in brackets). A few remarks about different elements in this table are, however, essential in this connection. NEON AND META-NEON. NEBULIUM. Evidence for the existence of' two isotopes' of neon has recently been deduced by Prof. J. J. Thomson and Aston. By careful diffusion experiments the latter was able to separate from neon another gas of atomic weight 22,. which has been named meta-neon. The two gases differ only in their gravitational • properties, but are chemically and spectroscopically identical. During the past year spectroscopic evidence has been adduced for the existence of a new element netraliti, having an atomic weight of about 3. This element occurs in the spectrum of the nebula of Orion. it is, however, probably too premature to try to speculate about its place in the Periodic Table. There are a number of elements like nebulium for the existence of which we have. only spectroscopic evidence, and it may be, as has been suggested recently, that these are the protoelements out of which our terrestrial elements have been built up. RARE EARTHS The case of the rare earths has already been discussed in a previous- section. The arrangement shown in Fig. 4 is in accordance with the atomic numbers determined by Moseley in the case of the following elements: Lanthanum, cerium, prreseodymium, neodymi1im, samarium, europium, gadolinium and holmium.- The order of atomic numbers in the case of dysoprosium and holmium is apparently the reverse of that of the atomic weights. But this case, as well as those of tellurium, iodine; cobalt, nickel; and argon, potassium, no longer appears anomalous when the elements are arranged in order of increasing atomic number rather than- -'that of increasing atomic weight; The atomic weight of neoytterbium has been ' determined during the past year; “it is, however, impossible to state at present what relation it bears to the” other' elements of the rare earth group. RADIOACTIVE ' ELEMENTS. The radioactive elements have been arranged in groups' of isotopes and the atomic numbers are based upon the order of the different elements in the disintegration series (see Fig. 3), assuming the atomic number of lead to be 82. The atomic weight of actinium and its disintegration products have not been determined. We have therefore adopted the value suggested by Fajans which is about 227. All we can say definitely is that the atomic weight'is greater than that of radium and considerably less than that of thorium. The atomic weights of uranium and radium are based on the' following considerations: First, as radium is derived from uranium by the expulsion of three alpha -particles, the atomic weights must differ by 3 X 3.99 units.-:- Secondly, according to the most recent report of -the International Oommittee on Atomic Weights -there seep! to be valid reasons for accepting a - value' which is very close to 238.2 for the atomic weight of uranium. The value actually obtained by Hoenigschmid (Z. Elect. 20, 452, 1914) varied from 238;W.:to 238.18; but the committee consider the latter.value -as being the more accurate. The determinations of- the atomic weight of radium have yielded results varying from 225.9 to 226.4, and the latter is the value given in the Table of Atomic Weights issued by the International. Committee for the present year. However, in wiew 0f. the above considerations we have used the value 226,2. ' The nomenclature of the radioactive elements is based on that of Soddy!, At the time when they were isolated, there was of,.course no definite knowledge as to their relationship and the. result has therefore been rather confusing... Thus the name polonium has been applied to RaF, while UX21s also known as brevium. The designation “niton” for radium emanation has become quite well knowa It has, however, been considered advisable to, use those- names which best convey the relationships of the different elements, and an attempt has been made to carry out this plan in tabulating the isotopes. CONCLUSIONS. Considering the relationships exhibited by the different radioactive. elements, one realizes that the dream of'the alchemists may not have been as fatuous as has appeared until recently. The concept of an absolutely stable atom must be discarded once for all, and its place is taken by this miniature solar system, as it were, consisting of a central nucleus and one or more rings of electrons. But the nucleus itseif is apparently the seat of fumense' forces, and in spite of its exceedingly. infinitesimal dimensions it contains both alpha particles' and electrons. Once in a while the nucleus of one of the atoms will spontaneously disintegrate and expel a.n alpha or beta particle. A new element has been born. What causes these transformations? Can they. be controlled? These are questions which only the future can answer. But if we had it in our power to remove two alpha particles from the atom of bismuth or any of its isotopes, not only would the dream of the alchemists be realized, but man would be in possession of. such intensely powerful sources of energy that all our coal mines, water-powers, and explosives would become insignificant by comparison. REFERENCES. 1. Pattison Muir—History of Chemical Theories and Laws, 2. F. Soddy—The Chemistry of the Radio-Elements, Parts I and II. 3. K. Fajans—Naturwissenschaften, Vol. II, 429, 462 (1914).