162 CHEMICAL DISCOVERY AND INVENTION 



of the elements and their relations to one another has grown 

 bolder. The discovery of the periodic relations of the elements 

 by Newlands, Mendeleefl, and Meyer may be regarded as the 

 starting-point of much of this modern hypothesis, though, 

 strangely enough, not shared by Mendeleeff, who devoted the 

 greater part of his life to the subject. 



The earliest serious treatment of the periodic relations of the 

 elements as the basis of ideas as to their origin we owe to Sir 

 William Crookes. His first exposition of his views was given in 

 1886 at the Birmingham meeting of the British Association, but 

 he has on several occasions revived the hypothesis then brought 

 forward, with modifications suggested by successive discoveries 

 already described. The last formal utterance of his views was 

 addressed to the International Chemical Congress at the meeting 

 in Berlin in 1903. Briefly his ideas on " The Genesis of the 

 Elements " are based on the assumption of a " protyle," as he 

 calls it, in the form of a mist of minute particles, which gradually 

 accreting into larger and larger clusters gave rise to elemental 

 atoms, which became assorted by a selective process de- 

 pendent on the tendency of particles with the same kind and 

 rate of motion to separate from a crowd and keep together. In 

 order to explain the production of the chemical elements, as 

 known, it must be supposed that in this process of accretion 

 some clusters of particles are more stable than others, and 

 therefore survive the changes of conditions (temperature, pressure, 

 etc. ?) to which they are exposed throughout the ages which 

 have elapsed since they were formed. To do justice to his views 

 it is necessary to quote Crookes' own words, taken from the 

 Proceedings of the Royal Society (vol. 63, p. 409. 1898) in 

 which he describes a model designed to represent diagram- 

 matically the evolution of the elements. 



" I take any arbitrary and convenient figure of eight without 

 reference to its exact nature ; I divide each of the loops into 

 eight equal parts, and then drop from these points ordinates 

 corresponding to the atomic weights of the first cycle of elements. 

 In the model the elements are supposed to follow one another 

 at equal distances along the figure of eight spiral, a gap of one 

 division being left at the point of crossing. The vertical height 

 is divided into 240 equal parts on which the atomic weights are 

 plotted from H = l to Ur= 239-59. l Each black disc represents 



Since corrected to 238 '2. 



