TEANSACTIONS OF SECTION B. 635 



abled twenty-five years ago. If to this he added tlie fact that foreign nations are 

 becoming increasingly hostile and exclusive commercially, we cannot feel surprise 

 at the dismal forebodings entertained and the contident predictions of decline 

 uttered by some who claim to know all the facts. 1 ought to ajjologise for alluding 

 to so gloomy a subject on the occasion of this to a great extent festive gathering, 

 but then men of science like to look at a question not only from a hopeful but from 

 every point of view. Fortunately on this question they are not called upon to 

 pronounce any opinion one way or the other. 



Should this be the last time that Manchester shall entertain the British Asso- 

 ciation in the day of its prosperity, I can only say with the German poet — 



Schliesst den Kreis und leert die Flaschen 



Diese Sommernilchte feiernd, 

 Schiimme Zeiten werden kommen, 



Die wir audi sodann ertragen. 



Whether in prosperity or adversity I feel sure that this city will always endea- 

 vour to entertain its visitors to the best of its ability. On the present occasion I 

 may, with confidence on the part of the chemical world of Manchester, offer to the 

 many friends from near and far who honour us with their presence at this meeting 

 a most hearty welcome. 



6. Freliminary Notice of a Re-determination of the Atomic Weight of Gold, 

 tfith some remarJcs on the fresent State of oti,r KnoivUdge as to the 

 Determination of Atomic Weights in general. Bij Professor J. "VY. 

 Mallet, F.R.S. 



For the last two years experiments had been in progress looking to as exact a 

 determination as possib!.e of the atomic weight of gold, for which until recently 

 there had been but very few data. Within the last few months the results had 

 been published of two researches on this subject by others, namely, by Kriiss in 

 Germany, and. by Thorpe and Laurie in England. Yet Mv. IMaliet's work was 

 continued, since there can scarcely be too much verification of important constants, 

 and the methods adopted were not altogether the same as those used by the other 

 chemists named, all whose results were obtained by essentially one and the same 

 process. Such liability to error as belongs to this process was pointed out. 



The authoi-^s own work had not yet reached the point of giving final results for 

 publication, but the probability seemed to be that a rather higher value would 

 be found for the atomic weight in question than tliat assigned by other 

 experimenters. 



It was suggested that the most important direction for advance in our know- 

 ledge of atomic weights is that of endeavouring to eliminate ' cmistant errois,' as 

 distinguished from mere personal or casual errors of experiment. The latter we 

 have been taught, by the example of Stas, to reduce to very small values by 

 minute and elaborate precautions. But the former are always to be suspected, and 

 all conceivable means should be used to avoid them. Among the most important 

 of such means the following were pointed out : — 



1st. — Resort in every case for the purification of materials used to ' fractional ' 

 methods, assuming materials to be pure only when earlier and later fractions give 

 sensibly identical results. 



2nd. — Great care in the study of the reactions depended upon for final determina- 

 tion of atomic weiglits, looking especially to any possibility of the occurrence of 

 secondary or subsidiary reactions. 



3rd. — Adoption of methods by which («) the atomic weight to be determined 

 may be connected directly with that of hydrogen, or (b) if connected indirectlj', 

 by the intervention in each single determination of as feiu other elements, but iu 

 determinations by diflerent methods of as many other elements as possible of 

 supposed well-known atomic weight. 



Under this last head a method was described which had been resorted to by the 

 author in his work on the atomic weight of gold, affbrdiug a direct connection with 



