440 
PROFESSOR J. W. MALLET OR THE ATOMIC WEIGHT OP GOLD. 
The results of the fifth and sixth series, obtained by electrolysis, are, I am convinced, 
much less entitled to confidence than any of the others. If these two be excluded, 
the general mean of the remaining series will be 196’882, a number differing but 
little from the mean of all. 
The highest value is that derived from the fourth series—ignition of trimethyl- 
ammonium auri-chloride. It has been seen that the individual results of this series 
agree fairly well with one another, and, when examined in connection with the facts as 
to the different crops of crystals of the salt used, do not seem to present any evidence 
of want of uniformity in the materiai. But, as it may still be suspected that traces of 
dimethyl- or of monomethyl-ammonium auri-chloride may have been present, and have 
caused the apparent value of the atomic weight of gold to come out higher than the 
truth, if we exclude also this series, the general mean of the remaining four ^rill 
be 196-796. 
Finally, if for the sake of comparison with the results of Kru ss and of Thorpe and 
Laurie the general mean be taken for the first three series only, in which auric 
chloride and bromide were examined, the result is 196-762 —intermediate between 
the general means of the two previous researches, but rather nearer to that derived 
from the work of Thorpe and Laurie than of Kruss. 
It will be observed that, although there is pretty close agreement among the means 
of results obtained by altogether different methods, this agreement is not so close as 
that presented by the results of the nearly similar methods pursued in the first three 
series. This cannot but suggest the probability of there being still sources of minute 
errors inherent in the methods themselves, and not dependent upon mere imperfec¬ 
tions of manipulation in carrying these methods out. Although there is thus to be 
noticed a slight tendency on the part of each method to yield high or low figures 
severally, with the exception of the results of the fourth series there does not appear 
to be any considerable reason to see in the values ol)tained confirmation of the special 
suspicions in connection with each method which have been stated. There is no clear 
evidence of any difference in the results which can be traced to the history of the 
particular samples of gold used ; a larger number of somewhat low results seem to 
have been yielded by the metal designated as (C)— i.e., obtained from the United 
States Assay Office at New York—than by the others, but the difference is not 
marked or constant enough to warrant any trustworthy conclusions as to the character 
of this material. 
Conchidmg Remarks. 
The atomic weight of gold as deduced from the experiments reported in this paper 
is entirely in accord with the place occupied by the metal in Mendelejeee’s periodic ” 
classification of the elements, but this is equally true of the slightly different values 
obtained by Kruss, and by Thorpe and Laurie, and the only difficulty at one time 
apparent as to this point—namely, the relative positions of gold on the one hand and 
