THE TESTING OF THE LAW. 35 



not only very like but they actually are the predicted ele- 

 ments eka-boron, eka-aluminum and eka-silicon respec- 

 tively. How unquestionable this fact is may be readily 

 seen from, for example, a comparison of MendeleerFs eka- 

 boron with the scandium afterwards discovered by Nilson. 



Eka-boron. Scandium. 



Atomic weight 44. Atomic weight 44. 



Oxide Eb 2 3 - Oxide Sc 2 3 . 



Sulphate less soluble than Sulphate less soluble than 



aluminum sulphate. aluminum sulphate. 



Carbonate insoluble in water. 'Carbonate insoluble in water. 

 Salts colorless and form ge- Salts colorless and form ge- 



Lcinous precipitates with latinous precipitates with 



potash and ammonia. potash and ammonia. 



The predicted properties are identical with the properties 

 found. The hypothetical element discovers itself in scan- 

 dium. 



In like manner the other two elements discovered them- 

 selves in gallium and germanium. We see in this success- 

 ful three-fold prediction the scope and power of the periodic 

 law as an instrument of research. We see convincingly 

 that the law must be the expression of a fact. Suppose 

 that an astrologer informed you that your horoscope led 

 him to believe that you would meet, sometime in your life, 

 three men; and that with the utmost particularity he told 

 you their weights, the color of their hair, the size of their 

 noses, and, in a word, all the habits of mind and body suffi- 

 cient to differentiate them positively from all other men ; and 

 suppose, moreover, that you met these men possessed of 

 qualities identical with the description predicted. You 

 would believe in astrology. 



Astrology cannot do these things, but chemistry can 



