I THE general laws of affinity are independent of the qualitative difference of matter; they have been discovered and applied without the aid of the atomic hypothesis and the latter indeed might appear to be superfluous for the representation of the most essential portion of chemical phenomena. The atomic hypothesis, in fact, did not seem indispensable for the explanation of the phenomena of chemical affinity, so long as the stoicheiometric laws alone confirmed it, and so long as atoms were considered to be the elementary, ultimate and inalterable particles of matter. But at the present time the question has assumed an entirely different aspect; certain ingenious researches which have been made in some branches of physics and the discovery of the radio-active bodies have lent to the theories of the granular constitution of matter the new and un for seen support of experimental confirmation; in short, we are forced to believe in the actual existence of atoms and of molecules as a logical consequence of all the facts known to us. Furthermore, the atom can no longer be considered as the simplest primordial particle of matter but rather as an element of matter having a highly complex internal structure; for this reason the problem of affinity presents itself to us in an entirely different fashion: .Hitherto it was possible to make an abstraction of the atomic hypothesis and this is what was done in fact when we confined ourselves to general laws such as are derived from the application of the second principle of thermodynamics; today, since we must regard atoms as physical units of which _ we cannot make an abstraction, it seems indispensable, ' on the contrary, to regard them as actually and essentially concerned in chemical phenomena. The latter depend upon the intimate constitution of the atoms, in which we naturally find the origin of the qualitative differences of matter connected with chemical forces. Affinity is a form of energy which, according to fundamental laws, can be transformed directly and re'versibly into heat or into electromagnetic energy. The relations which exist between chemical and thermic energy are derived from the principles of thermodynamics ; it will be of value to investigate the significance which these relations may acquire If we adopt modern views concerning the constitution of atoms. According to the kinetic theory all movement of atoms ceases at the point of absolute zero, at which point their calorific capacity seems, therefore, to be anulled. The internal energy of atoms, on the contrary, does not disappear at extremely low temperatures, as is demonstrated by the phenomena exhibited by radioactive bodies and by the experiments of Moissan concerning solid hydrogen and fluorine which still react violently. If we accept the kinetic theory it Is not difficult to understand that chemical actions, measured by the rapidity of the reaction, will be proportional to the degree of concentration and will be accelerated by elevations of temperature. But as was proved by Van't Hoff, the temperature also exerts an effect upon chemical equilibrium, especially in cases of dissociation. At this point we may enquire what is represented by the absorption of heat in these reactions as well as in endothermic processes in general. It is evident ' that this absorption of heat measures the transformation of the external movement of atoms in movements of their intimate constituents, or, to give a general expression, to this thesis, in internal atomic energy, which energy-is also capable, of course, of assuming the potential form. If we did not assume this variation of the internal energy of atoms in endothermic compounds, it would become necessary to suppose that the formation of bonds between atoms occasions an expenditure of energy and this would not agree with the various electromagnetic representations which may be conceived for atoms themselves. In our opinion the formation of bonds between atoms occasions a development of energy in endothermic compounds as well as in exothermic compounds; but in the case of the former this energy would not be sufficient to compensate for the expenditure required to augment the internal energy of the atoms which are combining. It results from this that the algebraic sum of the energies brought into play In these combinations is negative. According to the kinetic theory, heat results from the kinetic force Iforce vivej of atoms or of molecules in vibration or in translation: II reversible process in which the chemical energy which Is transformed into thermic energy' may be represented by a mechanism according to which the progressive or vibratory movement of atoms is transformed In a reversible manner into potential energy or into internal movement of their constituents. in other words we may say that the free energy of atoms is the fraction of their internal energy which is capable of becoming external in the form of heat or in any other form without the decomposition of the atom. II. If we assume, as it seems necessary to do, that chemical affinity is a function of the internal energy of atoms, it would be interesting to seek the determining cause in a more precise manner. The most diverse hypotheses have been formulated along this line, but we shall not attempt to analyze them here. We may assume not only the existence of electrostatic attraction, which the atom is capable of either by the loss or the reception of electrons, but also of electro-static attraction between neutral atoms, if we adopt the ingenious ideas of Rutherford' and of Bohr! These authors suppose that an electron revolving around an atom with a small positive mass in its center, is capable of exerting upon another positive ion or atom' electro-static actions which are opposite and dependent upon the distance at which it is situated. This viewpoint is also sustained by Crehore, who has sought to demonstrate how with such a scheme it is possible to construct formulas, especially of organic compounds.' Finally, to mention only the principal hypotheses, Ramsey* has attempted in his last work to explain chemical actions by the electromagnetic attractions which result from the reciprocal action of the circuits formed by the orbits of electrons and which may differ according to the direction of the rotation around the atom. It is our opinion, furthermore, that to comprehend the essential part of the chemical phenomenon it is not sufficient to take into consideration electrostatic or electromagnetic attractions alone; whatever the theory one may formulate with respect to the nature of chemical affinity, we believe it is necessary to consider a coefficient which must be especially made manifest even though it may be comprised implicitly-& some one of the hypotheses which we have mentioned above. This coefficient is related to the character of the chemical phenomenon which has always served to define it as a whole and which resides in the change— sometimes slight but often very great—of the properties of a compound with respect to those of the components. When we examine the matter minutely it seems hard to believe that two bodies, no matter how small they may be—two atoms, for example—can undergo an essential change of their properties merely because they are placed in new conditions of reciprocal attraction. The chemical phenomenon is of very special nature and cannot reside solely In an exchange of attractions. This constitutes a weak point which has enabled the adversaries of the atomistic theory to reject the ancient hypothesis of Dalton because it could not explain the differences which exist between the properties of free elements and those of their compounds. Today, thanks to the famous experiments of Bragg we know that the atoms in crystals are merely juxtaposed; in a crystal of sodium chloride, for example, the two kinds of atoms alternate in the crystallographic network and the molecules mingle with each other in a regular succession of atoms of sodium and of chloride. What would be the difference between the crystalline sodium chloride and a mixture in which the same number of atoms of sodium and of chlorine were arranged with the same regularity? It seems to our mind, therefore, logical to believe that the essence of the chemical phenomenon is a modification of the internal structure of the atom; this modification which the atom undergoes at the moment of combination may involve either a loss or a .gain of energy. The atoms in the molecules are juxtaposed, 'Phil. Mag., 26: 476, 857; 19: 332; J0: 394. 'Phil. Mag., 1911, U: 669: !7: 488. 'Phil. Mag., »0: 613. 'Proo., ROI/al Society, London, Series A, II!: Ml. indeed, but their internal structure is no longer the same as in the free atom." These modifications of the internal structure may determine attractions of electrostatic or electrodynamic origin; and if it is difficult to reconcile the two facts that we can always extract an element again from a compound and that an atom does not appear to exist in a combination with all its properties, the difficulty disappears if we imagine that the structure of the atom is modified at the moment of the combination, and that it regains its original state when the element becomes free once more. For the present we have no knowledge of the real nature of these changes. The different hypotheses which have been formulated in recent years concerning the possible structure of atoms are certainly of great value, but are still far from being able to solve the problem. Nevertheless, the studies of J. J. Thomson • and of Crehore', the experiments made by G. Newton-Lewis', and above all, the marvelous experiments and deductions of Moseley', blaze the way towards such a solution.10 As things stand at present, and in order to take into account the nature of these transformations of the internal structure of atoms at the moment of combination, it may be useful to compare them to a certain exent, to metamorphoses of organic compound such as those of the indicators [indicateurs] in salification. We now know that in this case the indicator is not salified by simple ionization, remaining unaltered meanwhile, but that its internal structure undergoes a change and is re-established when the indicator returns to its original state” by means of the inverse operation. In the same way the expulsion or the acquisition of an electron at the moment of combination, must be accompanied by a more or less profound internal transformation of the atom; this enables us to comprehend the actual difference which exists between the atom and the corresponding ion, a difference which the mere presence of an electric charge does not suffice to explain, in spite of all that has been said •A change of the Internal structure of atoms In combination explains naturally and convincingly why the properties of a compound may be different from those of the elements which It contains. Nevertheless, the following objection to this method of regarding the question might be preferred: while certain properties of the atom change when It enters Into combination others persist, as If the atom existed unaltered In Us compound. The specific heat In the solid state, the refractive power, etc., have an “additive” character, and this is particularly true of radio-activity, which Is not Influenced by the chemical combinations. But In the consideration of the last mentioned property we may find an argument with which to make reply to the objection offered above. As a matter of fact, temperature, which exerts a very great influence upon the state of combination, exerts none at all upon radio-activity, at any rate within the bounds of our present knowledge. Furthermore, the quantities of energy brought into play In ordinary combinations are much Inferior to those which are inherent in the phenomena of radio-activity. We may hold the opinion, therefore, that the transformation of the atom which we suppose to be accomplished when it enters into ' combination does not involve those constituents of the atom which determine the physical properties mentioned above, and we may hold likewise that a modification of these constituents would compromise the existence of the atom. By way of a comparative example It may be recalled that in the case of certain organic compounds a change of structure slightly modifies both physical and chemical properties: these compounds include the open-chain bodies and the corresponding alicycllc compounds, the parafflnes and the cyclo-parafflnes. We may also call attention to the heats of combustion which are almost equal for certain isomers. 'Phil. Mag., 1914, 26 : 792, 1044 ; !7: 747. 'Phil. Mag., 26: 25; 19: 310. "Jour. the Am. Soc., 1916, 762. 'Phil. Mag., 1914; !I: 1624; !7: 703. "An excellent resume of this question has been prepared by W. P. Harkins and E. D. Wilson, Jour. Amer. Chem. Soc., 1915, 1396. "Apropos of this it may 'be recalled that colorless phenol-phtalein for example, changes first of all its internal structure when its molecule comes in contact with a base R.OH when becoming transformed into an acid having a violet-red color, which afterwards combines with the base in becoming salified. If a stronger acid H.N be afterwards added the original phenol-pthalein will be reconstituted, but we do not have merely a simple double exchange in this Instance for the colored acid which separates out becomes transformed by an Internal metamorphosis Into the colorless original pseudo-acid. In the same way when di-methyl-amido-azo-benzol, which Is a yelow base is salified with an acid H.N, the salification determines a change of structure, since the compound of the quinonic type is more stable; the salt which Is formed is no longer nitric, 'but quinonic and of a red color. By adding a stronger base we may occasion the Inverse reaction, which will regenerate the yellow nitric compound. September 6, 1919 SCIENTIFIC AMERICAN SUPPLEMENT No. 2279 155 upon this subject. We can now readily conceive that an atom (or a molecule) can acquire an electric charge without becoming thereby chemically different from an electrically neutral atom, just as, for example, ionized air does not differ from non-conducting air. The atom becomes an ion by losing or gaining energy according to the nature of the internal transformation which it undergoes (independently of the hydration which may accompany the phenomenon) and it will remain more or less stable itself according to the greater or less degree of stability present in the new state. This greater or smaller stability of the internal constitution of the ion with respect to that of the atom corresponds to what has been rather vaguely called the intensity of adhesion of the ion to its charge. The greater or lesser tendency of the atom to acquire the structure of the ion corresponds in its turn to what is called the tension of the electric solution of the elements which is expressed numerically in the electrochemical series. Furthermore, the variation of structure in an atom which becomes an ion—accompanied according to our hypothesis by a variation of internal energy—may possibly explain the origin of the heat of ionization, particularly the positive heat which it is difficult otherwise to comprehend. Ill. The nature and physical significance of valence has been the subject of the most diverse hypotheses. The theory of dissociation caused that of valence to assume a new aspect: according to the law of Faraday the valence of an atom is determined by the number of electrons which it may lose or acquire. Among those who have helped to throw light upon this concept is J. J. Thomson, who supposes that the attractive actions determined by the valences are exerted not in a spherical field, as in the case of other attractions but solely in the direction of certain lines or channels of force to which he now attributes a cylindrical form; these channels of force through which atoms exert the attractions of valence may be regarded as corresponding in some sort to the marks (short straight lines) by which chemists represent the units of valence, without, however, giving to them any special physical signification. Making an abstraction of the nature of these attractions, which, as we have said above, may be of different origin, we believe that our idea may be expressed by saying that the number of affinities at the disposal of an element in its combinations, depends upon what may be termed the external form of the atom. We have arrived at this conclusion by studying the results of the researches made by Bragg. As we know, the crystallographic net-work of the diamond” is so constituted that each atom of carbon is connected with four other atoms along lines running from the center to the apices of a tetrahedron. This fact corresponds in a most surprising manner to the tetrahedral arrangement of the valences of the carbon atom, to which chemists have had recourse to explain certain isomeries of organic compounds; we think it logical, therefore, to believe that carbon functions as a tetra-valent element in its compounds, since its atom assumes a tetrahedral form. But this does not imply, obviously, that the free atom will have the same form. We believe this manner of interpreting the valence of atoms can be generalized and applied to the other elements. But if we accept the hypothesis we are confronted by two possible interpretations: must we believe that the form of atoms varies in the different types of combination, or assume that it remains constant ? The question may be considered from the following point of view: Certain elements retain the same chemical habit throughout the different types of combination. For example, oxide of carbon, hydrocyanic acid and the carbylamines, tri-phenyl-methane and the more complex analogous compounds do not differ essentially from the corresponding compounds of carbon in which the atom acts as tetravalent By analogy we may say . that the oxygenated compounds of chlorine possess common chemical characteristics In spite of the different amounts of oxygen contained, and the same thing is true of the compounds of nitrogen. On the other hand certain other elements undergo so great a change of character In passing from one type of combination to another that one might think that the same atom consists of two or more different elements. Thus the knowledge of cupric salts could not enable us to pre diet the existence of the cuprous salts; and how re markable does thallium appear because of the great difference which exists between the thallous and the thalllc compounds I The same thing is true of the compounds of manganese, if we compare the manganic and the permanganic acids with the manganous salts; or in the compounds of chromium, the derivatives of chromic acid with the salts of the sesqul-oxide. When we take into account such facts as these, it seems necessary to believe that in the case of some ele ments the form of the atom may vary according to the type of the combination and consequently that there are elements whose atoms are polymorphous. In the case of some of them, such as carbon, for example, the variation of the valence does not imply the modification of the form of the atom, for the valences remain free; in the case of others different forms of the atom correspond to different types of combinations; for example, the atom of copper Is dimorphous. In the case of additional compounds when complex ions are formed according to the law of Werner, the atoms and the entire molecules which enter into play constitute new polyhedrons with greater or lesser variations . of energy and of internal constitution. The chemical value of these new polyhedrons Is given, as we know, by the dierence between the valence of the element and those of the co-ordinated groups. We cannot make an absolute distinction between the double salts and the complex salts, but there is a gradual passage from one to the other. It does not seem to us to be necessary or even useful to suppose that the atomic attractions are of diverse origin in order to explain the existence of compounds of this type. It would be preferable to say that affinity has always the same origin, but that its intensity diminishes in proportion as the atomic or molecular complexes become greater. The Wernerian stereo-isomeries may properly serve to demonstrate the idea that the valences of complexes are of the same nature as those of atoms. We can then compare a Wernerian complex to a large atom and it might happen that it existed in the free state if we adopt the hypotheses formulated for ammonium and the substituted ammoniums. As for the a-valent compounds (neutrons) such as tri-nitro-tri-amino-cobalt Co (No.) *• we can compare them to the atoms of the indifferent gases of the argon series. Modern concepts of the structure of crystals continually make more manifest the relations which must exist between isomorphism and the form of atoms and of molecules. It can be at once understood from what has just been said why certain elements are capable of yielding different series of isomorphic derivatives according to the type of combination; thus at the atom of thallium, for example, will have in its monovalent compounds a form similar to that of the alkaline metals in their salts, and likewise similar to that of aluminum in trivalent compounds. The type of combination often determines the relations of isomorphism between elee ments of different character, because of the fact that their atoms are themselves polymorphous. Finally, it should also be observed that in the relations between the radio-active elements and their latest products of disaggregation there have recently been discovered facts of great importance which agree with the views set forth above; two or more atoms having the same structure and the same form may be identical in their properties, even if the quantity of matter which they contain is slightly different (isotopic ele ments) and reciprocally two elements may be different in spite of possessing an identical atomic weight, i. e. when containing equal quantities of matter if their structure and their form are different Hence the properties of elements are not a function of their atomic weight” alone as was believed by Mendeleeff but also of their intimate structure.Saving by Skip-Stops Data are given from tests made with gasoline and electric cars concerning the energy-saving made possible by “skip-stops.” The statement made by Government engineers that 10 to 16 per cent saving could be made by skip-stop operation is confirmed by these tests. The percentage increase in power consumption on comparative stops was as follows: 1 Car Car | Car Car A B C D Stops every 600 ft., over no 1 89% 187% 960/0 24% Stops every 300 ft., over no 148 213 167 66 Stops every 300' ft., over 600 81 32 37 33 Illuminating Devices in the Great War (Continued from page 147) temperature is determined. The following calculation adapted from Lissak's “Ordnance and Gunnery” will show how these values are obtained. Notation: Qmp = Heat given off by a molugram of the mixture at constant pressure and surrounding temperature t. That is, the difference in heat between the right and left hand side of the equation. Qmv = Heat, etc., as above at constant volume. N, = Number of unit volumes after expansion to normal atmospheric pressure and O°C. For hea t developed we have, (Qmv = Qmp + 0.572 N" (1) Then for one kilogram of the mixture or substanejg..e hare Qtotal = X (2) Total Mol. Wt. ' For temperature obtained we have Qmv = ti, (3) Where C mv = Molecular specific heat in small calories. t. = Rise in temperature in degrees Centigrade. Equation (3) would hold true if the specific heat of the products were constant. This is not quite true however as the specific heat increases with the temperature according to the following law, . °mv = a + Mi, (4) If we assume 15°Cent. as the initial temperature, then t = t, + 15, (5) The following table gives values for the constants « and b for a few gases: Cmv 6.26 + 0.0037 t. 4.80 + 0.0006 t. Values of a are the molecular heats of the gases in small calories; values of b are the increments of the mean molecular heats for each degree rise in temperature. By combining equations (3) and (4) and multiply-Inging Q mv by 1,000, since Qmv is in small calories, Gas. a II CO. and SO, 6.26 0.0037 H,O 11.61 0.0033 Gases Without condensation 4.80 0.006 we have, 1,00 = at, + bt", (6) This equation is a quadratic and when we solve for t we obtain, -° + v' o - 4,000 bJ Qmv -2b-"~ +15, t Cars A and B were double-truck cars, weighing 47,060 lbs. and 89,520 lbs. respectively. Car C was a single-truck car, weighing 28,740 lbs., and car D, a — passenger automobile, weighing 3,320 lbs. empty.— 11. .T. BvrnUck, tn Electric JRtMtooy Journal where t equals maximum theoretical temperature in degrees centigrade produced when the mixture burns. As this method of temperature and heat determination may be a little confusing to the reader at first it will be well to give an illustration. Let us take the same mixture which we used above, viz., barium nitrate and magnesium in the proportion of 68 to 32. The chemical equation of this reaction will be given here again for the reader's ready reference. Ba(NO.). + 5Mg = BaO + 5 MgO + N.. The heats of formation of these substances are given in the following table: Barium nitrate ......... 229 Large calories per molugram Magnesium oxide (MgO), .. 143 “ “ Barium oxide (BaO), ... 133 “ Specific heats are: Magnesium oxide.......10 Barium oxide...........10 Nitrogen................ 4.8 The heat equation may now be written as follows, 229 + 0 = 5 X 143 + 133 + 0, and the difference between the right and left hand side of this equation will be found to be 619 calories. In addition to this there is 0.572 large calories saved per molecular volume (22.4 liters) by the nitrogen not expanding. As there is but one part of nitrogen in the above reaction we have for the total heat, Qv = 619 + 0.572 X 1 = 619.6, and then Qtotal = (6!9-6 X 100) (262 + 121) = 1620, large calories per kilogram, at constant volume. The theoretical temperature of this same reaction may be obtained by substituting the known values in equation (7), thus: t= ~ 64-8+ V (64.8)' — 4,000 X 0.0006 X 619.6 2 X 0.0006 + 15 = 8840° C. The values of a and b were obtained in the following manner: a = a, + a. + a, ... = 5 + 10 + 10+ 4.8 = M.8 6 = b, + b. + b,.... = 0 + 0 + 0.0006 = 0.008 (Continued on pops 159) General view of the new electric locomotive of the Swiss Federal Railways. These engines now haul the trains of the Thun-Lotschberg-Simplon lines and the Bevers-Filisur section