on the Atomic Weights of Bodies. 261 



The first three of these numbers approach nearer the 

 mean quantity 0*375 than mine. But the difference is not 

 sufficient to warrant the conclusion, that Berzelius's num- 

 bers are right and mine wrong. Because a very slight 

 change in the number representing the specific heat would 

 bring the products of the atomic weights of arsenic, iron 

 and zinc as I have given them to the mean number 0-375. 

 Were the specific heats as follow : 



Arsenic, . . 0-079 instead of 0'081 

 Iron, . . . 0-107 „ 0-11 



Zinc, . . . 0-0909 „ 0-0927 

 the products of these numbers by the atomic weights would 

 be 0-375. Now, these deviations do not much exceed 1 per 

 cent., an error that may be easily committed in determin- 

 ing the specific heat of bodies. 



In my original experiments to determine the atom of zinc 

 I obtained 4*2. Now, the number 4- 125 which I have adopted 

 here comes almost as near 4*2 as my former number 4-25. 

 And I have preferred it, because, when multiplied by the 

 specific heat of zinc, the product approaches nearer to 0-375 

 than when 4-25 is taken. 



As to hydrogen, the number obtained by multiplying its 

 atomic weight by its specific heat deviates considerably from 

 0-375. For 0-412 = 0-375 + 37 or T Vth of the whole. The 

 reason of this I take to be, that it is much more difficult to 

 determine the specific heat of hydrogen gas with precision 

 than that of any other substance. And there are strong 

 reasons for believing that the number 3-2936 obtained by 

 Delaroche and Berard does not constitute a very near ap- 

 proximation to the truth. Had the specific heat of hydro- 

 gen been only 3 instead of 3*2936 its product multiplied by 

 0*125 would have given the mean number 0*375. 



It is obvious, at a glance, that the atomic weight of hy- 

 drogen adopted by Berzelius is too small. For, when 

 multiplied by the specific heat, the product does not much 

 exceed one half of 0-375. 



When we look at the product of the atom of azote 

 multiplied by its specific heat, it is obvious, that it is 

 about -J-th part too high. This I ascribe to the difficulty of 

 determining the specific heat of this gas with accuracy. 



