DALTON'S ATOMIC THEORY Not only were Black and Cavendish clearly convinced that chemical combination took place in definite proportions by weight, but all Lavoisier's work shows his belief in this law. On the other hand, Bergman and his followers had worked on the quan tities of acid and base required to form neutral salts, and Richter and Fischer had shown that there was a "reciprocal" law govern ing the weights of acids and bases which would just saturate each other. The several weights of different bases required to neutralize a given acid were also the weights required to neutralize a constant but different weight of another acid. This law of neutrality, embodied later in the "Law of Reciprocal Proportions," received little recognition at the time : it was revived to give sup port to and be explained by the Atomic Theory.
Dalton first briefly announced his theory in a memoir on the solubility of gases in 1803. He was led back to the old hypothesis of atoms through his study of the physical behaviour of gases, for he states that the solubility of a gas depends upon the weight and number of the ultimate particles of the gas. Already in 1801 he had shown experimentally that different gases under the same pressure expanded equally by heat, and had concluded that "all elastic fluids" did the same ; this fact he attributed to the ex pansion of a gas being entirely due to repulsion between the separated particles caused by heat alone, independently of any chemical attraction between them such as is shown in liquids and solids. His laboratory note-books show that Dalton had the idea of atoms in his mind some time before 1803, and the first table as published (in I 8o5) differs from the earliest tables found in his notes, but the essential facts are the same—if the weight of an atom of hydrogen is 1 a definite relative weight may be assigned to each atom of another element and to each compound atom formed by the union of these elementary atoms.
Dalton's table of the relative weights of the ultimate particles of gaseous and other bodies Hydrogen . . I Nitrous gas . 9.3 Carbonic acid . Azote . . . 4.2 Ether . . . 9.6 Alcohol . 15.1 Carbon . . 4.3 Gaseous oxide of Sulphureous Ammonia . . 5.2 carbon . 9.8 acid . 19.9 Oxygen . . 5.5 Nitrous oxide . 13.7 Sulphuric acid . Water . . 6.5 Sulphur. 14.4 Carburetted Phosphorus . 7.2 Nitric acid . 15.2 hydrogen . 6.3 Phosphuretted Sulphuretted Olefiant gas . hydrogen . 8.2 hydrogen • The Law of Multiple Proportion is clearly embodied in this table (in spite of mistakes in the weights given for nitrous "gas" and "oxide"). The two compounds of carbon with oxygen, the two compounds of sulphur with oxygen, and the two compounds of carbon with hydrogen are represented by the sum of their atomic weights—CO and CO2, SO and and CH.
Taking, as Dalton did, a fixed weight of one element in two or more compounds formed by it with another, then the weights of the second element which combine with the fixed weight of the first are simple multiples of each other. Dalton seems to have sought for instances of this chemical law to support the theory he had conceived on other grounds. Dalton published the full theory in his New System of Chemistry (18o8), representing his atoms by circles-0 for hydrogen, 0 for nitrogen, • for carbon, 0 for oxygen, etc., and placing these symbols to touch one an other in compounds; e.g., 00 for water, 00 olefiant gas, 0.0 marsh gas.
Dalton laid down certain rules for determining atomic weights, remarking that one must always presume where only one corn pound is known that it is a binary one, and that such compound should always be specifically heavier than the mixture of its two ingredients.
Dalton's theory was soon adopted, and elaborate researches were made—especially by Berzelius—to determine by analysis the correct atomic weights. One of the first changes in Dalton's weights was due to Davy's decomposition of soda and potash by the electric current, and the recognition that the alkaline earths were also oxides of metals.
In 18o8 Gay Lussac announced his law that gases when- they combine do so in very simple proportions by volume, and the product formed, when gaseous, also occupies a volume simply related to those of its ingredients. Hydrogen unites with chlorine in equal volumes, and the acid gas formed occupies the same volume as the two ; similarly two volumes of hydrogen unite with one volume of oxygen to form two volumes of steam measured under the same conditions. In such facts Gay Lussac saw a strong confirmation of Dalton's theory : assuming that the atoms of his gases occupied the same space, the volumes in which the gases combined ought to be simply related. But the idea that different atoms occupied the same space was repugnant to Dalton, who pointed out that steam was lighter than oxygen, and there fore in forming steam, according to Gay Lussac's idea, the oxygen atom must have divided itself.
The difficulty was overcome by Avogadro's Hypothesis (181i) that the chemical particles occupying the same gaseous volume were not single atoms but molecules consisting of 2 atoms, and in the formation of steam the oxygen molecule had divided itself between two hydrogen molecules. That "equal volumes of all gases, simple or compound,. contain the same number of molecules" was in agreement both with physical and chemical data. This generalization, however, was long opposed by Berzelius, who could not reconcile the diatomic molecule of a simple element with his electrochemical theory: moreover, the vapour densities of ele ments such as sulphur and phosphorus appeared anomalous.
With regard to the atomic weights of the metals the law of Dulong and Petit (1819) that "the specific heats are inversely proportional to the atomic weights," and the law of Mitscherlich (182o) that "salts having the same crystalline forms have similar chemical constitution?' were of great assistance : but in spite of much detailed progress great confusion existed in the use of dif ferent "equivalents" and atomic weights. Order was only restored by the reform advocated by Cannizzaro (1858), who insisted that the atomic weight deduced from the weight of the gaseous molecule agreed with that calculated from the specific heat, and should be adopted.
Meanwhile, the elements had been grouped into families some of which showed marked resemblances between their members. Moreover, it was observed that the members of a family showed a regular gradation in their chemical and physical properties and in their atomic weights. Dumas drew attention to several such "triad" groups and showed that the central element was nearly the mean, in weight and properties, between the first and third ; for instance sodium was half-way between lithium and potassium ; bromine half-way between chlorine and iodine ; selenium half-way between sulphur and tellurium ; strontium half way between calcium and barium. Elements seemed to be cast in certain moulds. Similarly, the formulae of compounds were built up on certain simple types. The use of the water-type by Williamson and of the ammonia-type and methane-type by Hof mann did much to explain similarities in chemical reactions, and led to the recognition of the different valencies of the elements. That atoms had valencies or bonds by which they attached them selves to each other, and that one atom differed from another in this power, gave a reason for the several types—if it was assumed that elements like hydrogen had one valency, others like oxygen had two, and so on. Frankland, who introduced this conception, accounted for elements having varying valencies by supposing that two of the bonds might be mutually satisfied, so that a pentad atom like =N = (with five bonds) might also act like a triad by the self-junction of two bonds,