on the Atomic Weight of Carbon. 215 



were multiples of the equivalent of hydrogen. It is in this 

 manner that the number 75 is arrived at as the atomic weight 

 of carbon, that number being very nearly the multiple of the 

 atomic weight of hydrogen, by the whole number 12. 



This is doubtless an unusual mode of controlling the accu- 

 racy of an experiment, yet the numbers thus modified have 

 been admitted by many chemists. 



The fact that the atomic weights of simple bodies are very 

 nearly the multiples of that of hydrogen by whole numbers, 

 is in itself nothing remarkable; there are other numerical 

 relations of this kind, which appear far more extraordinary, 

 on the strength of which, however, no one would think of 

 making a change in atomic weights. 



Thus, if we add to the atomic weight of potassium '189*92 

 the atomic weight of lithium, whicli is 80*33 



we obtain the number 570*25 



and this divided by 2, gives 285*12; now this last is as near 

 the atomic weight of sodium, which in chemical properties is 

 a link between the two former bodies, as is the atomic weight 

 of that body obtained by multiplying that of hydrogen by a 

 whole number. 



Again, the sum of the atomic weights of barium and of 

 calcium, divided by 2, gives very nearly that of strontium; 

 the sum of those of chlorine and of iodine, divided by 2, gives 

 nearly that of bromine; and those of iron and cobalt, divided 

 by 2, give that of manganese. 



There are evidently hidden connexions in these numerical 

 relations, with which we are not acquainted, and to take them 

 as criteria, before they are understood, is obviously incon- 

 sistent with the true spirit of philosophy; the same must be 

 admitted with respect to the hypothetical atomic weight of 

 carbon, 75, for the correctness of which as yet no experience 

 speaks. The early determination of the atomic weight of 

 lead by Berzelius, entirely confirmed as it is by his later ex- 

 periments in 1830, is in complete contradiction with the cor- 

 rectness of atomic weights as multiples of that of hydrogen. 



His memoir on that subject is indeed highly important in 

 reference to our own investigation, and should not be for- 

 gotten by chemists. The following is an extract verbatim 

 from his memoir*. 



"Further experiments on the Atomic Weight of Lead and its 

 Oxides. — The reduction of a metallic oxide by means of 

 hydrogen gas, appears so simple an experiment for determi- 



* Poggendorff's Annakn, B. xix. S. 310-315. 



