BETWEEN THE ATOMIC -WEIGHTS. 243 



the series ; but in others it presents remarkable variations, owing probably to the fact 

 that it contains a radical which is easily decomposed. As well known, it is perfectly 

 isomorphous with Chlorine. 



The propriety of classing Oxygen in this series seems to be placed beyond doubt by 

 the discovery of Ozone, which, though it does not seem to possess such energy as we 

 should expect in an element higher in the series than Fluorine, may, nevertheless, be 

 found to fulfil all anticipations should it ever be obtained in a perfectly unmixed con- 

 dition. The isomorphism of Oxygen with Chlorine, and therefore with the other ha- 

 loids, seems sufficiently established by the determination both of Proust and Misterlich 

 of the tetrahedral form of CuaCl. It must, however, be admitted that Oxygen presents 

 as strong analogies with Sulphur as it does with Chlorine ; and since, not only from its 

 analogies, but also from its atomic weight, it appears to be the nucleus in all the first 

 three series, I have placed it at the head of each. It may be mentioned here, that in * 

 all cases the fact that the atomic weight of an element is included in the general formula 

 of a series, is an argument for classifying it in that series, if the relation between the 

 atomic weights pointed out in this memoir is admitted to be a law of nature ; but as I 

 wish to show that the relation is not that of a mere accidental group of numbers, but 

 is connected with the most fundamental properties of the elements, and has, therefore, 

 the claims of a law, I have endeavored to establish the correctness of the classification 

 which confonns to the law, and, indeed, suggested the law on other grounds. 



The atomic weights of the numbers of the Nine Series, as determined by experiment, 

 present greater deviations from the numerical law already explained, than are to be 

 found in any of the others. The weights which would exactly conform to the general 

 formula 8 -\- n9 are given in the column of the table headed Theoretical, while in the 

 next column at the right are given the weights of experiment. These for the most 

 part (in this as well as in the other series) have been taken from the table of Atomic 

 Weights given in the last volume of Liebig and Kopp's Jahresbericht (for 1852), 

 which was supposed to give the most accurate and latest results. In the few cases in 

 which the numbers have not been taken from this table, the initial letter of the name 

 of the observer has been annexed. It will be seen, on comparing the two columns 

 that the greatest deviation from the law is in the case of Fluorine, if we consider the 

 care which was taken both by Berzelius and Louyet in the determination of the atomic 

 weight of this element. It may, however, be remarked, that, as the processes used by 

 both experimenters were essentially identical, any hidden constant source of error 

 would produce the same efiect on both results ; so that the atomic weight of Fluorine 

 cannot be regarded as yet as absolutely fixed. Nevertheless, it is not possible that so 



VOL. V. NEW SERIES. 3-4 



