20 SCIENCE PROGRESS 



will contain i8( ==2 x 3^) cells, and the fourth 32( =2 x 4^) 

 cells. These cells will all be arranged symmetrically with 

 reference to some diametrical plane ; this appears to be the signi- 

 ficance of the factor 2. The successive atomic types are then 

 constructed by adding successive electrons to each cell, and of 

 course adding corresponding positive unit charges or protons 

 to the nucleus. Thus in hydrogen one of the inner cells contains 

 an electron, in helium the two are occupied. In lithium a third 

 electron is added to one of the eight cells in the second shell ; 

 beryllium has two of these cells occupied, boron three, carbon 

 four, and so on until fluorine is reached with seven of them 

 occupied and then neon with all the cells in this shell occupied by 

 one electron each. We have then completed the first short period 

 of the table. The second short period from sodium to argon 

 corresponds to a further addition of one electron at a time still 

 to the second shell, and the next inert gas, argon, is reached 

 when the second layer of this shell is occupied completely by 

 eight more electrons than neon. There now follow two long 

 periods of 1 8 elements each in the table, and these correspond to 

 successive occupation by electrons of first the 1 8 cells in the first 

 layer of the third shell, and then the 1 8 cells in the second layer 

 of the third shell ; and when a whole layer of cells is occupied 

 we arrive at more inert gases, kr3^pton and xenon. Finally 

 there is the last complete period of 32 carrying us up to niton ; 

 but only a few elements between niton and uranium correspond- 

 ing to the occupation of a few cells in the second layer of the 

 fourth shell. 



The great power of this model is the really remarkable way 

 it accounts for the properties of the elements and compounds, 

 especially the facts of valency, which the older theories of val- 

 ence were quite inadequate to deal with. (For further details 

 consult the references at the end of this note.) But while it is 

 possible to suggest certain principles of statics which account 

 for the maintenance of electrons in equilibrium in such posi- 

 tions as are suggested, it is very difficult to admit that such 

 equilibrium would be stable under the considerable disturbing 

 influences to which every atom is subject, i.e. if stability is to be 

 interpreted in terms of the older dynamics. Further, if the 

 physical properties (especially optical) are to be adequately 

 explained by such models, it is necessary to permit of oscillations 

 or rotations of the molecules around the postulated equilibrium 

 positions, and, as mentioned above, this involves a general 

 scheme of dynamics which is so far wanting. Now, although a 

 general scheme is still to be sought, a particular principle due 

 to Bohr and based on the quantum hypothesis has been applied 

 with some success to a number of individual cases of electronic 

 orbits ; but as a general rule the models postulated in these 



