320 Dr. T. Carnelley on the 



W. C. Williams and T. Carnelley (Chem. Journ. 1879, 

 p. 563) have shown that a curious relation (whether accidental 

 or not it is impossible to say) exists between the melting- and 

 the boiling-points of CI, Br, I, and of S, Se, and Te. The 

 melting-points of the latter elements are respectively twice 

 and the boiling-points three times as high as those of the first- 

 named elements, all being reckoned from the absolute zero 

 (-273°), thus:— 



Melting-points. 



CI = 198 X 2 = 306 (Berthelot) S = 388 (Person). 



Br = 248x2 = 406 (Baumhauer)* Se = 490 (Ilittorf). 



I = 387 X 2 = 774 (Stas) Te = 773 (Watts's Dictionary). 



Boiling-points. 

 Cl= 240x3= 720 (Regnault) S =720 (Kegnault). 



Br = {331X3= 993( ( indEs)} Se = 953 (Williams and Carnelley). 



I = 473x3=1419 (Stas) Te boils below 1660, since Deville and 



Troost took the vapour-density of 

 Te at this temperature. 



The melting-point of I is equal to that of S, and double that 

 of CI. That of Te is double that of S. Similar relations also 

 hold in the case of the boiling-points, except that the boiling- 

 point of I is two thirds instead of equal to that of S. If the 

 boiling-point of I be twice that of CI ( = 240° x 2 = 480°), then 

 the number 473 is 7° too low ; but Stas gives the boiling- 

 point of I as slightly above (473-273 = 200°), thus confirm- 

 ing the calculated number. 



Applications of the Periodic Laic. — MendeljefF has pointed 

 out several important applications of the periodic law, of w T hich 

 the following are the more important : — 



(1) To the Classification of the Elements. — This is in fact 

 the only scientific and natural classification ; for it is the 

 only one which takes into account not only the chemical pro- 

 perties and atomicity, but likewise the physical properties and 

 atomic Aveights. 



(2) To the Determination of the Atomic Weights of Rare 

 Elements. — This will be best explained if we take an ex- 

 ample in illustration, such as indium. Having found the 

 equivalent of the element ( = 37*6), we calculate what its 

 atomic weight would be supposing it were a monad, dyad, 

 triad, and tetrad respectively, and we obtain the following 

 results : — 



Monad. Dyad. Triad. Tetrad. 



Atomic weight = 37*6 75-2 112-8 150-4 

 Now it cannot be a monad, because there is no vacant place 



* Baumhauer states (Deut. chem. Ges. Ber. iv. p. 927) that Br always 

 gives a melting-point two or three degrees too high if it is not perfectly 

 drv. 



