A.—MATHEMATICAL AND PHYSICAL SCIENCES 27 
the ordinary gas discharge which requires the element to be volatile or 
form suitable volatile compounds ; and the anode ray discharge, in which 
the halide or other compound of the element is treated as the anode in 
a discharge at low pressure. The inert gases are particularly suitable to 
the first method, the alkali metals to the second, other groups of elements 
being intermediate. Our knowledge of the mechanism of the discharge 
in both methods is far from complete, so that working with them is still 
rather an art than ascience. The element of luck has played an important 
part in cases where the properties of the materials are unfamiliar and 
unfavourable to the conditions of the discharge. 
The analysis of the recently discovered element rhenium offers a good 
example. The only available volatile compound was the heptoxide, 
a sample of which has been kindly provided by the discoverer Noddack. 
The vapour of this crystalline solid was first admitted to the discharge 
bulb, but without success. The solid was then introduced into the bulb 
itself, and, although its vaporisation was so copious that a visible layer 
was formed on the walls, still no lines were obtained. At this stage the 
element was abandoned as quite hopeless and preparations were made to 
goon to another. Purely by chance, this happened to be gold, which it 
was intended to attack by means of its slightly volatile chloride. This 
compound gives off chlorine gas when heated, and, as previously it had 
been noticed that the presence of a halogen gas often stimulated the 
appearance of lines otherwise faint, it was considered just worth while to 
make one trial with it before the rhenium oxide deposit had been cleaned 
off the walls. This was successful beyond all hopes. No lines of gold 
were found but the rhenium doublet appeared in great strength giving 
convincing evidence that it consisted of two isotopes, 185 and 187. 
The technique of anode rays is, if anything, even more capricious but, 
when successful, yields spectra almost free from the lines of compounds 
and is for this reason particularly suitable for the identification of new 
isotopes. This method has been recently applied to the large group of 
the rare earth elements yielding some thirty new isotopes. 
From the point of view of the identification of the more abundant 
isotopes our knowledge is nearly complete. A year ago only four elements, 
palladium, iridium, platinum, and gold, remained without mass- 
spectograph data. Dempster has since developed an entirely new method 
of obtaining suitable rays by using a very intense spark discharge, and 
I have just heard from him that he has already identified five isotopes of 
platinum and one of gold. It seems very probable that the last two 
elements will have yielded before this address is delivered. 
In all some 253 stable isotopes are known of which seven were dis- 
covered by observations on optical spectra, and have since been confirmed 
by the mass-spectograph. This large assembly shows many emprical 
laws, of which perhaps the more remarkable is that no odd numbered 
element, with the possible extremely rare exception the isotope of hydrogen 
of mass 3, has more than two isotopes. Even elements are not so limited. 
The most complex element so far observed is tin, with eleven isotopes 
ranging in mass number from 112 to 124. One of the most astonishing 
results is that, for practically every natural number up to 210, a stable 
elementary atom is known, many are filled twice over and a few three 
