PROFESSOR J. W. MALLET ON TITE ATOAOC WEIGHT OF OOliD. 
441. 
of platinum, iridium, and osmium on the other—lias been removed, not hy any cliange in 
the atomic weight of gold, but by changes affecting the values to be assiofiied the 
three other metals, as these values have been determined by Seubert.''' It is very 
desirable that, in order to a fuller and more exact examination of the Mendelejeff 
table of the elements, there be accomplished as soon as possible a general revision of 
the atomic weights of all the elements of well determined individuality, so many of 
which are still very imperfectly known. 
As to any bearing of the results of the present paper on the so-called hypothesis of 
PROUT,t the general mean of all my results, or even the general mean with exclusion 
of the values obtained by electrolysis, approaches the integer number 197 rather more 
nearly than does the final number arrived at by Thorpe and Laurie, and still more 
nearly than does that considered by Kruss to express the final result of his experi¬ 
ments. If the results of the fourth series be also rejected, my general mean will be 
nearer the integer than is the Kruss number, but not quite so near as that of Thorpe 
and Laurie. I feel that somewhat greater confidence may be placed in my own work, 
simply on the ground of its involving the use of more completely different and inde¬ 
pendent methods—a principle which I believe to be of the first importance in any 
attempts at increased accuracy in the determination of atomic weights. 
At the same time, as has already been pointed out, this work seems to me to furnish 
some probable evidence that not all inherent defects of method have been eliminated. 
Whether or not such defects may exist to an extent sufficient to account for the 
remaining difference between the value obtained and the integer multiple of the atomic 
weight of hydrogen there does not seem to be ground on which to express a positive 
opinion. But this research does not supply any clear evidence contradictory of such 
a possibility. 
On this point, and generally on the attainment of what is sometimes rather too 
easily spoken of as the greatest possible accuracy in the determination of an atomic 
weight—particularly of an element for which the value is as high as that for gold— 
any one who actually works in a conscientious way at such determinations will be 
pretty sure to feel more strongly the difficulty of the task, and to express himself 
with more caution, than do some compilers of results in assuming at any time that 
the last word has been spoken. 
* ‘ Bericlite Deatscli. Cliem. Gesell.,’ vol. 11, p. 1770; vol. 14, p. 868 ; vol. 21, p. 1869. 
t Soon after the publication of my paj^er on the atomic weight of aluminum, I wiis criticiseil by a 
writer of abstracts for the German Chemical Society on account of my use of the expression “Pkoct’s 
law,” amazement being indicated that I should have called the “ hypothesis ” of Prout a law. If this 
writer had noticed my use of inverted commas, and still more what was said in the course of two or 
three pages of the paper, he would have seen that the use of the expression “ Prout’s law ” was by no 
means equivalent to assuming this to be “ a law of nature.” 
3 L 
M DCCCLXXXIX.—A. 
