Page:1902 Encyclopædia Britannica - Volume 26 - AUS-CHI.pdf/767

 CHEMISTRY Kriiss, Mallet, and Thorpe and Laurie have shown that the value long accepted for gold needs but slight correction. Again, although the values assigned to titanium ranged from 48 to 57, Mendeleeff gave the preference to the former number as placing titanium relatively to silicon in the same position that vanadium bears to phosphorus, and it was shown by Thorpe in 1885 that 48 is the correct value. Tellurium alone is regarded by some as an exception. Being a member of the sulphur family it should come between antimony and iodine, just as sulphur comes between phosphorus and chlorine, and consequently its atomic weight should be lower than that of iodine. This has not yet been found to be the case,1 but there are special difficulties attached to the determination of the atomic weight of this element, and there is reason to suppose that the determinations are affected with unperceived constant errors. At the time Mendeleeffs first essay was published in 1871, only the equivalents were known of a number of the elements, and it is therefore noteworthy that, solely from a consideration of their properties, he came to the conclusion that it was necessary to take as the atomic weights other multiples of the equivalents than those commonly adopted in the case of beryllium, yttrium, indium, lanthanum, cerium, didymium, thorium and uranium. Excepting in the case of lanthanum his proposals have all been justified. Besults such as these certainly justify his statement that the periodic law enables us to master the facts relating to atomic weights, whereas before it was formulated atomic weights were purely empirical numbers, which could never be directly tested by considering the values themselves but only by critically examining the methods by which they were ascertained. The work done by Stas, and those who have been fired by his example, in seeking to place the determination of atomic weights on the most exact basis possible, will be more appropriately referred to later. It is impossible to overlook the importance of the discovery of so many new elements as have been referred to; Sufficiency obviously we cannot be too cautious in any of the conclusions we may deduce as to the number periodic likely to exist, and especially in imposing any theory. narrow limit. Moreover, the discovery of elements like argon, helium, etc., destitute of positive chemical properties, opens out an altogether fresh field of view, and at the same time greatly increases the difficulty of the discussion. What positions are to be assigned to them in the periodic series ? On what basis are they to be classified 1 As all the elements previously known are chemically active substances, and as there is nothing to foreshadow the existence of inert elements, chemical considerations afford no help in such a case, and it is only possible to classify them on the basis of their atomic weights. Their molecular weights can be deduced from their densities, but what of their atomic weights—the data by which alone their position can be decided 1 Rayleigh and Ramsay have found the ratio of specific heat at constant pressure to that at constant volume in argon to be 1*66 :1, which is the theoretical ratio for a gas in which all energy imparted at constant volume is expended in effecting translatory motion; they therefore assume that, like mercury, argon is a monatomic element. Helium, according to Ramsay, for a similar reason is also monatomic ; in fact, all the companions of argon in air appear to exhibit “ monatomic ” behaviour. On this view the atomic weight of helium is about 4, whilst that of argon is about 40. The former could be placed without difficulty between

711

hydrogen and lithium, but as the value 40 belongs to calcium, a place cannot be found for the latter. The element most nearly resembling argon is nitrogen. As we know it in atmospheric air, this is by far the least active of the elements, and yet we are satisfied that atomic nitrogen is gifted with intense affinities. The inactivity of molecular nitrogen, N2, may be ascribed to the exhaustion consequent on the union of two atoms which attract and bind each other so firmly that little opportunity is left for them to exercise external influence. It is only necessary to apply and somewhat extend this view to find an explanation of the apparently absolute indifference manifested by helium, argon,2 and similar gases. None of the non-metallic elements are known as monatomic molecules under any ordinary conditions; in fact, only metallic elements affect a monatomic state except at very high temperatures, and as the known gaseous elements— hydrogen, oxygen, nitrogen, fluorine, chlorine—all occur as diatomic molecules, it is permissible provisionally to make the assumption that this is true of helium and argon. Helium then follows hydrogen, and argon fluorine, both being appropriately placed after elements of exceptional activity. A difficulty of another kind, to which Crookes has specially drawn attention, occurs in the case of the rare-earth group of elements, of which, as already indicated, besides the considerable number now known, there is reason to believe many more exist. The differences they exhibit, judged by chemical standards, are extraordinarily slight; often there is at most a minute difference in basicity or solubility which makes it possible to separate one from the other by a process of fractionation repeated perhaps many hundred times. Are such slight differences to be regarded as proof of chemical individuality ? Crookes thinks not. His views on ment% didymia have already been referred to. Another case is that of yttria. Starting from as definite a substance as possible, prepared in such a manner that it would be commonly regarded as pure, he has succeeded, by long-continued fractionation, in separating from it six distinct fractions, each of which affords a spectrum in which the dominant band is a different member of the set of bands in the phosphorescent spectrum of the original material. Crookes does not consider, however, that the yttria is in this way separated into distinct earths. To explain these results, as well as those obtained in other similar cases, he has introduced a novel conception—that of meta - elements: that is to say, he would modify and widen the conception of an element, and would substitute that of an elementary group composed of meta-elements. His conception of an element may be gathered from the following passage:— In defining an element, let us not take an external boundary, but an internal type. Let us say, e.g., the smallest ponderable quantity of yttrium is an assemblage of ultimate atoms almost infinitely more like each other than they are to the atom of any other approximating element. It does not necessarily follow that the atoms shall all be absolutely alike among themselves. The atomic weight which we ascribe to yttrium, therefore, merely represents a mean value around which the actual weights of the individual atoms of the “element” range within certain limits. But if my conjecture is tenable, could we separate atom from

2 In the very stable gases the atoms in the molecule are combined very firmly, and therefore little internal work is done on heating them ; in the less stable, the combination being looser, more work is done internally on heating. The ratio Cp/C0, on this assumption, should be greater for the more stable gases, as appears to be the case judging from the values given by Clausius :— O2 N2 H2 CI2 Br2 1-40 1-41 1-41 1-30 1-29 If helium and argon are He2 and A2, and the atoms are gifted with a 1 Steiner, however, by analysing tellurium diphenyl, TePh2, has very high degree of mutual affinity, the molecules might well be so stable that no internal work is done on heating the gases. arrived at the value 126'4 [Ber. deut. chem. Ges. 1901, p. 570).