displays in a graphic form the real character of the difference between acids, bases, and salts. If, in the water type, for instance, represented by H-O-H, one atom of the hydrogen is replaced by a positive radical, we have + R-O-H, or a base. If it be replaced by a negative radical we have -R-O-H, or an acid. If both atoms of hydrogen be replaced, one by a positive the other by a negative radical, we have the symbol -R-O+R, or a salt.

In the ammonia type this principle has still greater extension, several acids and bases being formed by the successive replacement of three atoms of hydrogen. There are thus formed what are known as acid, neutral and basic salts, according as a portion of the acid or basic hydrogen remains in the salt, or as it is completely removed. In the ammonia type some chemists have given distinctive names to these products to separate them from their analogues in the water type. Thus an ammonia acid is an amide, a base is an amine, and a salt is an alkalamide.

The theory of types embraced an immense number of compounds, both mineral and organic, throwing down the wall which custom had raised between mineral and organic chemistry, giving to the notation an admirable clearness, and leading the way to numerous discoveries by enabling chemists to perceive at a glance the bonds of relationship between known and possible compounds. Yet it failed to go to the root of things. It did not show what these types really represented, or why hydrogen, water, and ammonia should be chosen as the typical compounds rather than others. This difficulty has since been resolved, by an important extension of the theory. The chemical types refer to the results of a fundamental property of atoms called atomicity. We will proceed to explain its character and application.

We will not trace it, however, through all the details of its history. These, though interesting, are too extended for our space. The details of the theory itself, with some features of its application, are all that we can give. Chemists arrived at the theory through observation of the reactions of complicated organic compounds. It can be best delineated by commencing

where they ended, with the elements themselves. There is, indeed, much significance in the formation of the typical compounds. It might be asked why chlorine is only able to attach to itself one atom of hydrogen where oxygen requires two, nitrogen three, and carbon four. This distinction, which extends through all the reactions of these elements, must have some important meaning.

There are compounds in which carbon is united with less than four atoms of hydrogen. But in these cases its affinity for hydrogen is not satisfied. It can take other atoms until it has gained four. Beyond this point a single atom of carbon cannot go. This then is considered to be its point of saturation, the atoms of carbon having four times the combining power of those of hydrogen. The same rule holds good in the cases of water and ammonia, and from it we derive the idea that atoms of chlorine, oxygen, nitrogen, and carbon, are equivalent respectively to one, two, three, and four atoms of hydrogen. This peculiarity has been signified in chemical language by calling these elements respectively univalent, bivalent, trivalent, and quadrivalent, which terms display their hydrogen-fixing power.

But the relations given here are not confined to these compounds. The elements retain this power through all their relations. Thus if hydrogen be replaced in any compound by chlorine the exchange takes place atom for atom. But if it be replaced by oxygen, two atoms of hydrogen are lost for every atom of oxygen gained. Here we have again the fact that an atom of oxygen is equivalent to two of hydrogen. This relation of the atoms to each other is called by Hofmaun quantivalence, the atom-fixing power of any, element or radical being designated by Roman numerals or dashes placed over the atomic symbol.

It appertains not alone to the elements we have named, but to all the elements, these being separated into classes according to their quantivalence, which extends from one to six in the different elements. The symbol of quantivalence then simply serves to show how many atoms of hydrogen or its equivalents the atom of the element is capable of replacing

in a compound, or how many such atoms are required to satisfy its affinities. The relations it thus displays in the case of hydrogen are carried through all its reactions with other substances, so that we are forced to conclude that these are no accidental or unimportant results, but that they point out distinctive properties hidden in the natures of the elements, and bring us to the threshold of possible discoveries in chemistry deeper than mere combinations and reactions, to a hope of some time gaining information as to the innate characters of the elements themselves, and the producing causes of the phenomena of affinity.

The quantivalence of the chemical elements is by no means always the same. Under different conditions they often exhibit different atom-fixing powers. Each element, however, has a maximum power, which it never exceeds. This is called .its atomicity, and the elements are distinguished as monads, dyads, triads, etc., according to their degree of this power. Thus nitrogen is a pentad, though it is more commonly trivalent, and lead is a tetrad, though it is usually bivalent. The elements are usually classed, however, according to their prevailing quantivalence, even though they sometimes display a greater atomicity.

It was as late as 1858 that the theory of atomicity began to gain its full extension, In that year Kekulé first broached the idea of the tetratomic power of carbon, being led to it by the fact that in the simplest organic compounds one atom of carbon is always united with a sum of elements equivalent to four atoms of hydrogen.* Thus, in carbonic acid, CO2, the two atoms of oxygen are equivalent to four of hydrogen Now carbon can content itself with one atom of oxygen, as in carbonic oxide, but then its affinities are not satisfied. We perceive this in its power of taking up another atom of oxygen, forming carbonic acid, or two atoms of chlorine, forming phosgene gas.

The organic hydrocarbons are capable of being extended into series, known as homologous series, becoming gradually

*Annalen der Chemie und Pharmacie, cvi., 129 (1858).

more and more complicated, yet retaining a fixed type of formation, each higher term in the series being formed by successive additions of a fixed sum of atoms to a basic compound. We may instance the series of alcohols, which grow more and more complicated by such equal additions, and the corresponding series of ethers, each of which differs from its analogous alcohol by the loss of H2O, equal to one molecule of

water.

In the series of saturated hydrocarbons having for base marsh gas, or CH, the second term is not C'H3, as would at first be supposed, but CH, and the advance is by successive additions of CH2. It is an important question how this fact can be made to conform to the theory of atomicity. It was answered by Kekulé in the memoir above cited. He supposes that the two carbon atoms, in the second term of the series, are combined by means of an atomicity of each, there being but six atomicities left, which are satisfied by the six hydrogen atoms. In the third term are three carbon and eight hydrogen atoms. We may conceive these carbon atoms as placed end to end, the central one being soldered to each of the other by one of its atomicities. Thus it has but two left, while each of the others have three left, having employed but one each. There are thus eight atomicities free for exterior work, and these are satisfied by eight atoms of hydrogen. Thus the carbon atoms form a chain whose links are riveted by part of the force of combination.

If now we take the second term of this series, composed of six hydrogen and two carbon atoms, we may replace one atom of hydrogen by one of chlorine, forming a compound known as chloride of ethyl. It is still saturated, for the chlorine atom is of the same value as the hydrogen atom which has gone out. But suppose, on the contrary, that it be replaced by an oxygen atom? The latter has two atomicities, one of which attaches it to the carbon atom, while the other is free. This is the real constitution of what is known as a radical, a group of atoms all whose atomicities are not satisfied, and which, therefore, tend to combine with other substances. In the present case there is but one atomicity to be

satisfied. The radical, therefore, is equivalent to a monad, and can only combine with a monad, as with an atom of hydrogen. If this combination be made, we have the original atom of hydrogen replaced by the group HO. But this group is simply a molecule of water which has lost an atom of hydrogen. It is then a radical with monad power, and can replace an atom of the monad hydrogen. In the case given, the compound known as ethyl hydride is by this change converted into ethyl hydrate, or alcohol.

But suppose an atom of nitrogen had replaced the hydrogen atom in the basic compound. It would need but one atomicity for this purpose, and would have two free atomicities. It would thus form a dyad radical, being capable of absorbing two hydrogen atoms, or their equivalent. Such a case really occurs, forming the compound known as ethyla

mine.

This theory had been applied to the organic compounds before it received its final application to the elements. For instance, it was known that where alcohol unites with only one molecule of a monobasic acid to form a compound ether, glycerine takes up three molecules of such an acid to form a neutral fat. Glycerine thus represents a tribasic alcohol, containing three hydrogen atoms, replaceable by three compound radicals.* We have already seen how the dyad radical sulphuryl can replace two atoms of hydrogen in two molecules of water. If now we conceive of three molecules of water in which three atoms of hydrogen are replaced by the triad radical glyceryl, we have an idea of the formation of glycerine which leads us, by a knowledge of its type of formation, to an understanding of its reactions.

This conception of the monatomic and triatomic alcohols led to a belief in the possible existence of diatomic alcohols. Efforts were at once made to isolate such compounds, and with gratifying success, the diatomic alcohols known as glycols being discovered. This soon led to the discovery of a series of compounds, the glycolic analogues of the higher terms of the alcohol series.

Wurtz, Théorie des Combinasions Glycériques.