Page:Encyclopædia Britannica, Ninth Edition, v. 5.djvu/482

470 C H E M I S T R Y [ATOMIC WEIGHTS. density is determined, provided always that the temperature be not so high as to cause decomposition. But the rela tive proportions in which the elements combine, and the composition of compounds may usually be ascertained with very great in many cases with almost absolute accuracy by chemical analysis; and the determination of the density in the state of gas simply serves to prove which of the several multiple proportions, in which it is found that the elements combine together, is the true atomic weight. For instance, analysis shows that marsh gas is a compound of carbon and hydrogen exactly in the proportions of 1 part of the latter with 3 of the former ; that carbonic anhydride consists of carbon and oxygen in the proportion of 16 parts or an atom of the latter and 6 or 3 x 2 parts of carbon ; and carbonic oxide of the same elements, in the proportion, however, of 1 G of oxygen and 1 2 or 3 x 4 of carbon. The question, therefore, is, Which of the numbers, 3, G, or 12, represents the relative weight of the carbon atom; that is to say, is marsh gas a compound of an atom of hydrogen and an atom of carbon weighing 3, of two atoms of hydrogen and an atom of carbon weighing 6, or of four atoms of hydrogen and an atom of carbon weighing 12? The molecular weights of three such compounds would be respec tively 3 + 1 or 4, G + 2 or 8, and 12 + 4 or 16 ; and the corresponding theoretical densities referred to hydrogen 2, 4, and 8. Finding, however, by experiment that the density of marsh gas is, say, 7 5, we at once conclude that the atomic weight of carbon is 12 and not 6 or 3, because the observed density of marsh gas most closely accords with that required on this assumption. The difference between theory and experiment is practically seldom, if ever, so large as in this case, which is merely given as an illustration of the principle involved. The equivalent of an element that is to say, the amount of it which is capable of combining with or displacing one part by weight, or one atom, of hydrogen being known, its atomic weight is not absolutely fixed by the determina tion of the density in the state of gas of its compounds with other elements ; we are at most enabled to say from this that the atomic weight cannot exceed a certain value, for instance, that the atomic weight of chlorine cannot exceed 35 36, because all its compounds contain either this amount or some simple multiple of it in their molecules. It is nevertheless jwssible that 35 3 G is not the weight of one but of several atoms of chlorine ; the probability that 35 - 3G is the true atomic weight is enormously increased, however, as compound after compound is examined and found to contain 35 36 or some simple multiple of 35 36 parts of chlorine in its molecule. In the case of those elements of which stable volatile compounds have not been obtained, the study of their specific heats is of great importance, and moreover furnishes most important confirmation of the atomic weights deduced by the aid of Avogadro s hypothesis. To raise the temperature of equal weights of different sub stances the same number of degrees, from to 1 C. for instance, very different amounts of heat are required ; and on the other hand very different amounts of heat are given out when equal weights of different substances are cooled from the initial temperature t to a lower temperature t . Of all bodies except hydrogen water has the greatest capacity for heat, and is, therefore, adopted as the standard of reference, the number which expresses the amount of heat necessary to raise the temperature of a given weight of a body a certain number of degrees, or which is given out by it in cooling through a certain number of degrees, as compared with that required to raise the temperature of an equal weight of water the same number of degrees, being termed its specific heat. Thus, the specific heat of lithium is 9408 ; that is to say, to raise the temperature of a given Name of Element. Specific Heat Atomic Weight. Specific Heat Multiplied by Atomic Weight. Antimony 0508 122 6-19 Arsenic 0814 74 9 6-10 Bismutli 0308 207-5 6-39 Boron 5 11 5-5 Bromine 0843 79-75 6-72 Carbon 4589 11-97 5-49 Iodine 0541 126-53 6-84 Lead 0314 206-4 6-48 Mercury 0317 199-8 6-33 Molybdenum 0722 95-3 6-92 Osmium 0311 198-6 6-17 Phosphorus 174 30-96 5-39 Selenium 9745 79 5-86 Silicon 2029 28 5-68 Sulphur 171 31-98 5-47 Tellurium 0474 128 6-07 Tin 0562 117-8 6-62 Tungsten .... 0334 184 6-14 Zinc 0955 64-9 6-19 weight of lithium 1 gramme, for example from to 1 C. only requires 9408 of the heat necessary to raise the temperature of 1 gramme of water from to 1 C. The specific heats in the solid state of the various elements of which the atomic weights have been determined by Avogadro s hypothesis are given in the second column of the following table : 50 144 100 191 150 200 236 -279 606 C 440 985 458 On comparing the numbers in the fourth column of this table it will be seen that they vary within comparatively narrow limits ; and if certain of the elements are excepted, viz., boron, carbon, phosphorus, sulphur, silicon, and selenium, the agreement becomes much closer, the average product obtained by multiplying specific heat into atomic weight being about 6 3. From this it would appear that the specific heats of these elements are, at least approximately, inversely proportional to their atomic weights. From the observation of this relation in the case of only a small number of elements Dulong and Petit, in 1811, were led to infer that the atoms of all simple bodies have the same capacity for heat. The specific heat of a body varies, however, with the temperature ; an extreme instance of this is afforded by the elements carbon, boron, and silicon, as will be evident on inspecting the following table of the specific heat of carbon in the form of diamond at various temperatures : Temperature O e Specific Heat 095 Product of AtomicWt. and Specific Heat .. Hence, owing to the circumstance that the determina tions of specific heat have not been made at temperatures which are comparable for the different elements, there is no doubt that many of the results which have been obtained are defective ; but from Weber s recent researches it appears that in the case of the solid elements there is a point for each element, after which the increase in specific heat with increase of temperature is insignificant, and when this point is reached the product of specific heat into atomic weight the so-called atomic heat varies within compara tively narrow limits. These limits, according to Weber, are from 5 - 5 to 6 5, but it appears probable that the superior limit is slightly greater than this, and as a matter of fact the atomic heats of nearly all the elements are nearer 6 5 than 5 5, the latter number being characteristic of the so-called non-metallic elements, the atomic weights of which can be determined by the aid of Avogadro s hypothesis. In consequence of this relation between the specific heat of an element and its atomic weight, we can readily deter- 1-12 172 2-28 2-81 3 33 448

5-26 5-36 5-48