134 Professor Frederick Soddy [May 18, 



The determination of the atomic weight of chlorine in terms of 

 that of silver has reached now such a pitch of refinement that it 

 shonld be able to detect a difference in the end fractions of the atomic 

 weight of chlorine, if chlorine or hydrogen chloride were systemati- 

 cally subjected to diffusion. It is extremely desirable that such a 

 test of the homogeneity of this gas should be made in this way. 



Clearly a change must come in this class of work. It is not of 

 much use starting with stuff out of a bottle lal)elled " purissimum " 

 or " garantirt," and in determining to the highest possible degree of 

 accuracy the atomic weight of an element of unknown origin. The 

 great pioneers in the subject, like Berzelius, were masters of the 

 whole domain of inorganic chemistry, and knew the sources of 

 the elements in Nature first-hand. Their successors must revert to 

 their practice and go direct to Nature for their materials, must select 

 them carefully with due regard to what geology teaches as to their 

 age and history, and, before carrying out a single determination, they 

 must analyse their actual raw materials completely, and know exactly 

 what it is they are dealing with. Much of the work on the atomic 

 weight of lead from mixed minerals is useless, for failure to do this. 

 They must rely more on the agreement, or disagreement, of a great 

 variety of results by methods as different and for materials as 

 different as possible, rather than on the result of a single method 

 pushed to the limit of refinement, for an element provisionally 

 purified by a dealer from quite unknown materials. The precon- 

 ceived notion, that the results must necessarily agree if the work is 

 well done, must he replaced by a system of co-operation between the 

 workers of the world checking each other's results for the same 

 material. A year ago anyone bold enough to publish atomic weight 

 determinations, which were not up to the modern standards of 

 agreement among themselves, would have been regarded as having 

 mistaken his vocation. If these wider ideals are pursued, all the 

 labour that has been lavished in this field, and which now seems to 

 have been so largely wasted, may possibly bear fruit, and where the 

 newer methods fail, far below the narrow belt of elements which it 

 is possiljle to watch changing, the atomic weight worker may be able 

 to pick up the threads of the great story. No doubt it is writ in 

 full in the natural records preserved by rock and mineral, and the 

 evidence of the atomic weights may be able to carry to a triumphant 

 conclusion the course of elementary evolution, of which as yet only 

 an isolated chapter has been deciphered. 



The Structure of the Atom. 



The third line of recent advance, which does much to explain 

 the meaning of the isotopes and the Periodic Law, starts from 

 Sir Ernest Rutherford's nuclear theory of the atom, which is an 

 attempt to determine the nature of atomic structure, which again is 



