ADDRESS. 9 



1837 to seventy in 1887 (not including the twenty or more new elements 

 recently said to have been discovered by Kriiss and Nilson in certain rare 

 Scandinavian minerals), but the properties of these elements have b6en 

 studied, and are now known to us with a degree of precision then undreamt 

 of. So that relationships existing between these bodies which fifty years 

 ago were undiscernible are now clearly manifest, and it is to these relation- 

 ships that I would for a moment ask your attention. I have already stated 

 that Dalton measured the relative weights of the ultimate particles by 

 assuming hydrogen as the unit, and that Prout believed that on this 

 basis the atomic weights of all the other elements would be found to be 

 multiples of the atomic weight of hydrogen, thus indicating that an inti- 

 mate constitutional relation exists between hydrogen and all the other 

 elements. 



Since the days of Dalton and Prout the truth or otherwise of Prout's 

 law has been keenly contested by the most eminent chemists of all 

 countries. The inquiry is a purely experimental one, and only those who 

 have a special knowledge of the difficulties which surround such in- 

 quiries can form an idea of the amount of labour and self-sacrifice borne by 

 such men as Dumas, Stas, and Marignac in carrying out delicate researches 

 on the atomic weights of the elements. What is, then, the result of these 

 most laborious experiments ? It is that, whilst the atomic weights of the 

 elements are not exactly either multiples of the unit or of half the unit, 

 many of the numbers expressing most accurately the weight of the atom 

 approximate so closely to a multiple of that of hydrogen that we are con- 

 strained to admit that these approximations cannot be a mere matter of 

 chance, but that some reason must exist for them. What that reason is, 

 and why a close approximation and yet something short of absolute iden- 

 tity exists, is as yet hidden behind the veil ; but who is there that doubts 

 that when this Association celebrates its centenary this veil will have been 

 lifted and this occult but fundamental question of atomic philosophy 

 shall have been brought into the clear light of day ? 



But these are by no means all the relationships which modern science 

 has discovered with respect to the atoms of our chemical elements. So 

 long ago as 1829 Dobereiner pointed out that certain groups of elements 

 exist presenting in all their properties strongly marked family character- 

 istics, and this was afterwards extended and insisted upon by Dumas. We 

 find, for example, in the well-known group of chlorine, bromine, and 

 iodine, these resemblances well developed, accompanied moreover by 

 a proportional graduation in their chemical and physical properties. 

 Thus, to take the most important of all their characters, the atomic 

 weight of the middle term is the mean of the atomic weights of the two 

 extremes. But these groups of triads appeared to be unconnected in any 

 way with one another, nor did they seem to bear any relation to the far 

 larger number of the elements not exhibiting these peculiarities. 



Things remained in this condition until 1863, when Newlands threw 

 fresh light upon the subject showing a far-reaching series of relation- 



