2 Wilde, (7;/ //le Atoimc Weight of TcUur'iinii. 



posing the four series, only amount to 0.0046 of the actual 

 determinations. 



The atomic weights are also in much closer agreement with 

 expcrmiental results than is the fundamental law of atomic 

 heats formulated by Dulong and Petit for these same series. No 

 one doubts the general accuracy of this law, because it does not 

 hold good for carbon, boron and silicon, or to fractional f|uan- 

 tities throughout the whole number of the elements. Dalton's 

 law of chemical combination in definite and multiple propor- 

 tions was founded on approximations differing for the princi]:ial 

 elements more than thirty per cent, from later determinations,"* 

 and through the adoption of the atomic weights of Cannizzaro, 

 these differences are largely increased. 



I would also emphasise the fact, hitlierto ignored by 

 chemists, that as the atomic weights of the two positive series 

 of elements, Hn and Fl2n, are the products of the large multiple 

 numbers, 16, 23 and 24 respectively, correlated also by the 

 common differences 4 and 8 with the large multiple numbers 46 

 and 48 of the two negative series of elements, the exact multiple 

 proportions subsisting among these higher atomic weights have 

 an immensely greater validity in determining the question of 

 their being whole numbers of hydrogen, than when all the 

 equivalents were compared directly with the unit or half-unit of 

 hydrogen by Stas and the older chemists. 



I liaxe now the honour to bring before the Academic a new 

 argument in favour of the exact multiple proportions of the 

 atomic weights, which, while helpful to earnest students of the 

 natural sciences, will be a permanent check to the pretensions 

 of those chemists who set up their laboured approximations 

 of the atomic weights as the absolute truth of nature and the 

 measure of the power of future investigators. 



In the memoir referred to, Dumas formulated the proposi- 

 tion that " in three simple bodies of the same natural family, 

 the ^-quivalent of the intermediate body, is always half the sum 

 of the equivalents of the two extreme bodies." This proposition, 

 as will be evident, 'is the rigorous expression of the definite 

 and exact multiple proportions of the atomic weights. 



The first example of this law givt^n by Dumas is the triad 

 of sulphur, selenium and tellurium, with the old ecjuivalents, 16, 

 40, 64, e(]ual to 32, 80, 128 of the atomic weights of Cannizzaro. 

 Now, in the geometry of solids, we have a triad of numerical 

 proportions similar to those found in the atomic weights, since 

 a cone, sphere and cylinder, of equal diameter and altitude, 

 have the ratios of i, 2, 3, respectively, and the intermediate 

 body is half the sum of the two extreme bodies, as in the triad 

 of sulphur, selenium and tellurium. The mental attitude of 

 those chemists who make their determinations of the atomic 

 weights the absolute truth of nature, would therefore be strictly 



4. Dalton's New S\ stem of Chemical Philosophy, Vol. II. )3. 35^ 



(lcS27). 



