In general, it has been found that each product shows some distinctive chemical or physical behaviour which allows of its partial or complete separation from a mixture of other products. It must be remembered that in most cases the amount of radio active matter under examination is too small to detect by weight, but its presence is inferred from its characteristic radiations and rate of change. In some cases, a separation may be effected by ordinary chemical methods ; for example thorium X is separated from thorium by precipitation of thorium with ammonia. The Th X remains in the filtrate and is practically free from thorium. In other cases, a separation is effected by a separation of a metal in the solution of active matter. For example, polonium (radium F) always comes down with bismuth and may be separated by placing a bismuth plate in a solution. Radium C is separated from radium B by adding nickel filings to a solution of the two. Radium C is deposited on the nickel. In other cases, a partial separation may be effected by electrolysis or by differences in volatility when heated. For example, when radium A, B and C are deposited on a platinum plate, on heating the plate, radium B is volatilized and is deposited on any cold surface in the neighbourhood. A very striking method of separating certain products has been observed depending upon the recoil of an atom which breaks up with the expulsion of an a-particle. Some of the residual atoms acquire sufficient velocity in consequence of the ejection of an a-particle to escape and be deposited on bodies in the neighbourhood. This is especially marked in a low vacuum. This property was inde pendently investigated by Russ and Makower and by Hahn. The latter has shown that by means of the recoil, actinium C" may be obtained pure from the active deposit containing actinium A, B and C, for C emits a-rays, and actinium C" is driven from the plate by the recoil. In a similar way a new product, thorium C", has been isolated and radium B may be separated from radium A and C. The recoil method has proved of great value in settling whether an a-ray product is simple or complex.
While in the majority of cases the products break up either with the emission of a- or 0-particles, some products were at first thought not to emit any characteristic radiation and were called "rayless products." A closer investigation has shown that in most
cases a f3-radiation of an easily absorbed type is emitted. Actin ium is the only product in which no certain radiation has been detected but from other evidence there can be no doubt that j3 rays of slow speed are emitted. The presence and properties of a product emitting a very feeble radiation can be easily inferred if it is transformed into a product emitting a marked radiation, for the variation of activity of the latter affords a method of deter mining the amount of the parent product. In the table below a list is given of the remarkable series of transformations occur ring in uranium, thorium and actinium. The system of nomen clature is not ideal. In general, products of average life greater than one year have been given a distinctive name and are printed in italics. The successive products following the emanations are in all cases denoted by the letters A, B, C, etc. In some cases inter mediate products have been discovered after a system of nomen clature has been accepted and have had to be named to indicate as far as possible their positions in the series. As the radioactive emanations are now known to be isotopic elements, the names radon, thoron and actinon have been suggested as distinctive and indicative of the similarity of these products.
Branch Products.—In the great majority of cases each of the ,adioactive elements breaks up in a definite way, giving rise to one a- or 0-particle and to one atom of the new product. Un doubted evidence, however, has been obtained that in a few cases the atoms break up in two or more distinct ways, giving rise to two or more products characterised by different radioactive prop erties. A branching of the uranium series was early demanded in order to account for the origin of actinium. While the latter is always found in uranium minerals in constant proportion with the uranium, Boltwood showed that the activity of the actinium with its whole series of a-ray products in a uranium mineral was much less than that given by a single a-ray product of the main radium series. The head of the actinium series is generally be lieved to be uranium Y, the branch product of the uranium series first observed by Antonoff. The branching is supposed to occur in the product uranium II, 3% going into the actinium branch and the other 97% into the main uranium series. The atomic weights of actinium given in the table are calculated on this basis. The more recently discovered product protoactinium is on this view the missing link between uranium Y and actinium.