We have also glanced at the changes of energy which accompany changes of composition, and we have seen that when equilibrium is attained in a chemical system there is a certain distribution of the members of the system, and also of the energy of the system.

We have arranged compounds in classes, such as acids, alkalis, salts, acidic oxides, &c. &c. We have found that all the members of each class shew certain similarities of composition. We have traced some of the connexions between the composition of each class and the typical property or properties of the class.

It has been found possible to assign to each member of certain classes of compounds a number, called the affinityconstant of the compound, which tells how much of a specified chemical change will be accomplished under definite conditions by that compound; and it has been found possible to establish the fact that there is a definite relation between this value and the composition of the compound, and, in some cases, to state the nature of this relation.

382 We have traced certain elements through the changes which they undergo by combining with other elements to form compounds belonging to different classes; we have seen the properties of the elements modified by the nature, and relative quantities, of the other elements with which they are combined; but we have not altogether lost sight of the original elements in these changes. Each element to some extent impresses its own properties on all the compounds of which it forms a part.

383

Is it possible then to connect the properties of the compounds of an element, using the term properties in its widest sense, with some one definite and measurable property of the element itself? Can it be shewn that the properties of the compounds vary with variations in the chosen property of the element?

If this can be done, we shall have the basis of a satisfactory method of chemical classification both of elements and compounds.

We have had instances of a regular variation of properties of similar compounds of a group of allied elements accompanying variations in the atomic weights of these elements (s. especially Chap. XI. par. 182).

Various chemists, among whom Newlands must be especially mentioned, from time to time have drawn attention

to points of connexion between the properties and the atomic weights of the elements.

VICHWED

In 1864 Newlands arranged a number of elements in order of their atomic weights, and shewed that these elements might be divided into groups of seven, each of which groups to some extent repeated the properties of the next group. "The eighth element said Newlands "starting from a given element is a kind of repetition of the first, like the eighth note of an octave in music." In subsequent papers Newlands insisted on the general applicability of what he called the 'law of octaves.'

It is however chiefly to Mendelejeff that we owe the systematic correlation of the atomic weights with the chemical and physical properties of the elements, and the properties of their compounds.

The properties of the elements and their compounds vary 384 periodically with variations in the atomic weights of the elements. This statement is the outcome of the work of Mendelejeff, Lothar Meyer, Newlands, and many other chemists.

This statement, or a statement equivalent to this, is known as the periodic law. The rest of this book will be devoted to an attempt to amplify and explain the periodic law.

The properties of the chemical elements and their com- 385 pounds is the phenomenon to be examined; the variable is the atomic weight of the elements: the law asserts that with a continuous change of the variable the phenomenon repeats itself at definite intervals.

A quantitative value cannot be given to the phenomenon 'properties of the elements and their compounds.' To illustrate the statement of the periodic law, it is necessary to choose a definite measurable property of the elements, or of a series of similar compounds, and to trace the relation between the variation in the values of this property and the variation in the values of the atomic weights of the elements.

We shall choose two properties, one physical, the other partly chemical and partly physical. The properties are (1) the melting points, (2) the atomic volumes, of the elements. The atomic volume of an element is defined to be

this quotient represents the volume occupied by a mass of the

solid element proportional to the atomic weight of the ele

ment.

The most striking way of exhibiting the connexion between the variable, atomic weight, and each of the variants, melting point and atomic volume, is to represent the values of the former as lengths marked off on a horizontal line, and the values of the latter as lengths on a vertical line; then, from each pair of points so marked to produce lines until they meet, and to draw a curve cutting the points of intersection of these lines. The curves thus obtained are drawn in the plate on page 269. The melting points are calculated from the initial temperature -273°; the numbers so obtained are divided by 7 in order to bring the curve within manageable limits. The values of the atomic volumes are multiplied by 4 to make the scale of the curve comparable with that of the curve of melting points. Lack of data is indicated by a broken line, or by a gap in the curve. Thus if the elements are arranged in order of increasing atomic weights, nitrogen, oxygen, and fluorine come after carbon and before sodium; but the atomic volumes of these elements are unknown, hence the dotted line in the curve of atomic volumes. Similarly a number of elements come between didymium and tantalum; the gap in the curve indicates that the atomic volumes of these elements have not been determined. The melting points of only about two-thirds of the elements have been determined; hence many parts of the curve of melting points are shewn as dotted lines.

386 The curves shew that the melting points, and the atomic volumes, of the elements vary periodically with variations in the atomic weights of the elements. The value of either variant does not exactly repeat itself at definite intervals; but the elements fall into periods, in each of which the values of the melting point and the atomic volume decrease from a maximum to a minimum, and then again increase to a maximum. The nature of this periodical variation is best shewn by the curve of atomic volumes, as the data are here more abundant. The first period comprises the elements from lithium to sodium; the second, the elements from sodium to potassium; the third, the elements from potassium to rubidium; the fourth, the elements from rubidium to caesium; after caesium there is a great want of data. Elements, the values of whose atomic weights place them about midway between the first and last element of a period, have atomic volumes

250

€ 240

Fe

Ru

Co

Pd

Cs

1

Ni

ATOMIC VOLUMES ARE MULTIPLIED BY 4. MELTING POINTS (Calculated from-273) ARE DIV? BY7.

1

230

220

210

200

Au

Bi

Di

TBI

Pb
T

150

Mg

140

130

120

110

Br

ATOMIC VOLUMES AND MELTING POINTS

100

Zn

90

80

70

Th

70 80 90 100 118 120 130 140 150 160 170 180 190 200 210 220 230 240 Thick line curve shews atomic volumes.

Thin

melting points.

Dotted lines indicate that data are wanting

THIS POINT SHOULD BE PLACED 66 DIVISIONS HIGHER

387

approximately equal to those of elements which occupy a similar position in another period; compare, for instance, the positions and atomic volumes of chromium, manganese, iron, nickel, and cobalt, with the positions and atomic volumes of rhodium, ruthenium, palladium, and silver; or compare the positions and atomic volumes of sulphur, selenion, and tellurium.

If the only properties of elements which it was necessary to study were their melting points and atomic volumes, it is evident that the connexion between the values of these and the values of the atomic weights of the elements is so marked and definite that a system of classification might well be based on this connexion. But the periodic law asserts that the properties of the elements and their compounds in general, and not only one or two properties in particular, vary periodically with variations in the atomic weights of the elements.

Let the 14 elements from lithium (Li=7) to chlorine (Cl = 35.5) be arranged in two series or periods of seven in each; thus—

=

Li = 7 Be 9. B=11 C=12 N=14 016 F=19. Na 23 Mg=24 Al=27 Si = 28 P = 31 S=32 Cl = 35.5. The difference in the values of the atomic weights of two consecutive elements varies from 1 to 3.5, the mean difference is about 2. The difference between the values of the atomic weights of two elements placed one under the other varies from 15 to 17; the mean difference is about 16.

The following statement gives a general indication of the chemical properties of lithium and sodium, beryllium and magnesium, boron and aluminium, &c.

Lithium and sodium: very light, soft, easily melted, metals; rapidly decompose cold water, thus

M+H2O = MOH + H;

oxides, MO, alkali-forming and strongly basic; form salts M.SO, M.CO, MCI, &c.; do not combine with H.

ვი

Beryllium and magnesium: fairly hard metals, of low spec. gravity but high melting points; oxides basic but not alkali-forming; form salts MSO,, MCO,, MCI,, &c.; do not combine with H.

Boron and aluminium: Al metallic, B non-metallic; B forms BH,, Al does not combine with H; Al forms salts