[Reprinted from the Journal of the American Chemical Society, 
Vol. XVII., No. 3. March, 1895.] 
REPORT OF COMMITTEE ON ATOHIC WEIGHTS, PUB- 
LISHED DURING 1894.' 
By F. W. Clarke. 
Received JanuaFFyTWy?. 
To the Members of the American Chemical Society: 
YOUR committee upon atomic weights respectfully submits 
the following report, which summarizes the work done in 
this department of chemistry during 1894. Although the volume 
of completed determinations is not large, it is known that several 
important investigations are in progress, from which valuable 
results may be expected in the near future. It is in this coun- 
try that the greatest activity exists, and that the greatest prog- 
ress is being made at present; and the preparation of these 
reports is therefore a peculiarly appropriate function of the 
Society. The data for 1894 are as follows: 
The H: O ratio.-An interesting attempt at the indirect meas- 
urement of this ratio, which is the base line upon which our sys- 
I Read at the Boston Meeting, December 28, 1894. 202 
F. W. CLARKE. REPORT OF COMMITTEE 
tern of atomic weights depends, has been made by Julius Thom- 
sen.1 His determinations are really determinations of the ratio 
NH3:HC1, and were conducted thus: First, pure, dry, gaseous 
hydrochloric acid was passed into a weighed absorption appara- 
tus containing pure distilled water. After noting the increase in 
weight, gaseous ammonia was passed through to slight excess, 
and the apparatus was weighed again. The excess of ammo- 
nia was then measured by titration with standard hydrochloric 
acid. In weighing, the apparatus was tared by another as nearly 
like it as possible, containing the same amount of water. Three 
sets of weighings were made, with apparatus of different size, 
and these Thomsen considers separately, giving the greatest 
weight to the experiments involving the largest masses of mate- 
HCl 
rial. The data are as follows, with the ratio in the third 
nh3 
column : 
Wt. HC1. 
wt. nh3. 
Ratio 
5.1624 
2.4120 
2.1403 
3-9425 
1.8409 
2.1416 
4-6544 
2.1739 
2.I4II 
3.9840 
1.8609 
2.1409 
5-3295 
2.4898 
2.1406 
4.2517 
1.9863 
2.1405 
4.8287 
2.2550 
2.1414 
6-4377 
3.0068 
2.1411 
4.1804 
1.9528 
2.I407 
5-0363 
2.3523 
2.14IO 
4.6408 
2.1685 
2.I4I I 
First Series. 
Second Series. 
Wt. HCE 
Wt. NHS. 
Ratio. 
11.8418 
5-5302 
2.14130 
14.3018 
6.6808 
2.14073 
12.1502 
5-6759 
2.14067 
H-5443 
5-3927 
2.14073 
12.3617 
5-7733 
2.14118 
Wt. HC1. 
Wt. NHS. 
Ratio. 
19-3455 
9.0360 
2.14094 
I9-4578 
9.0890 
2.14081 
Third Series. 
Ztschr. phys. Chern., 13,398. ON ATOMIC WEIGHTS, PUBLISHED DURING 1894. 
203 
From the sums of the weights Thomsen finds the ratio to be 
2.14087, or 2.13934 in vacuo. From this, using Ostwald's reduc- 
tion of Stas' data for the atomic weights of nitrogen and chlo- 
rine, he gets the ratio O : H : : 16 :0.99946, or almost exactly 16:1. 
In a later paper1 Thomsen himself recalculates Stas' data, with 
O - 16 as the basis of computation, and derives from them the 
subjoined values for the elements which Stas studied: 
Ag 107.9299 
Cl 35-4494 
Br 79-9510 
I 126.8556 
S 32.0606 
Pb 206.9042 
K 39-I5O7 
Na 23.0543 
Li 7-0307 
N 14.0396 
Combining these values for chlorine and nitrogen with his 
ratio HC1: NHS he gets O : H :: 16 :0.9992. This, however, is 
only an apparent support of Prout's hypothesis, for it depends 
upon the anti-Proutian determinations of Stas. If we calculate 
from Thomsen's new ratio with N = 14 and Cl = 35-5, it gives 
H = 1.0242 ; which is most unsatisfactory. In short, the method 
followed by Thomsen is too indirect and subject to too many 
possibilities of error to entitle it to much weight in fixing so 
important a constant as the atomic weight of oxygen. The 
direct processes, followed by several recent investigators, and 
giving 0= 15.87 to 15.89 are much more trustworthy. Meyer 
and Seubert5, in their criticism of Thomsen's work, have pointed 
out some of its uncertainties. 
In this connection it may be noted that Scott's research upon 
the composition of water by volume, cited by abstract in the 
report of last year, has been published in full in the Philosoph- 
ical Transactions.3 
Strontium.-The atomic weight of strontium has been rede- 
termined by Richards4 from analyses of the bromide. The first 
ratio measured, after a careful preliminary study of materials 
1 Ztschr. phys. Chem., 13, 726. 
2 Ber d. chem. Ges., ay, 2770. See also abstract by Ostwald in Ztschr.phys. Chern.,15, 705. 
8 184, 543,1893. 
4 Proc. Amer. Acad. 1894, 369. 
2-13-95 204 
F. w. CLARKE. REPORT OF COMMITTEE 
and methods, was that between silver and strontium bromide. Of 
this ratio, three sets of determinations were made, all volumet- 
ric, but with differences of detail in the process. The weights 
are as follows, with the ratio Ag2: SrBra:: 100 :x in the third 
column : 
First Series. 
Wt. Ag. 
Wt. SrBr2. 
Ratio. 
I-3O755 
1.49962 
114.689 
2.10351 
2.41225 
114.677 
2.23357 
2.56I53 
114-683 
5.36840 
6.15663 
114.683 
Sum, 11.01303 
12.63003 
114.683 
Wt. Ag. 
Wt. SrBr2. 
Ratio. 
1.30762 
I.49962 
II4.683 
2.10322 
2.41225 
II4-693 
4-57502 
5-24727 
114.694 
5.36800 
6.15663 
114.691 
Sum, 13.35386 
I5-3I577 
114.692 
Second Series. 
Wt. Ag. 
Wt. SrBra. 
Ratio. 
2-5434 
2.9172 
II4-697 
3-3957 
3.8946 
114.692 
3.9607 
4-5426 
114.692 
4-575° 
5-2473 
II4-695 
Sum, 14.4748 
16.6017 
114.694 
Third Series. 
From these data we have, if Ag = 107.93, and Br= 79.955, 
(O = 16), the following results : 
From first series Sr = 87.644 
" second series 87.663 
" third series 87.668 
In two additional series, partly identical with the foregoing, 
the silver bromide thrown down was collected and weighed. I 
subjoin the weighings with the ratio 2AgBr:SrBr2 in the last 
column. 
2AgBr. 
SrBr,. 
Ratio. 
2-4415 
I.6086 
65.886 
2.8561 
I.8817 
65.884 
6-9337 
4-568l 
65.883 
Sum, 12.2313 
8.0584 
65-8834 
First Series. ON ATOMIC WEIGHTS, PUBLISHED DURING 1894. 
205 
Second Series. 
2AgBr. 
SrBrs. 
Ratio. 
2.27625 
1.49962 
65.881 
3.66140 
2.41225 
65.883 
3-88776 
2-56I53 
65.887 
9-34497 
6.15663 
65.882 
Sum, 19.17038 
12.63003 
65.883 
From the first series Sr = 87.660 
" " second series 87.659 
The average of all five series is Sr = 87.659. 
Barium.-Richards has corroborated his earlier determinations 
of the atomic weight of barium, which were made with the bro- 
mide, by means of additional series of experiments upon the 
chloride.1 The work was carried out in the most elaborate and 
thorough manner, and for details the original paper must be con- 
sulted. First, barium chloride was titrated with standard solu- 
tions of silver, and the several series represent different methods 
of ascertaining accurately the end point. The data are as fol- 
lows, with the ratio Ag,2: BaCl2: : 100 : x in the third column. 
First Series. 
Wt. Ag. 
Wt. BaCla. 
Ratio. 
6.1872 
5-97I7 
96.517 
5.6580 
5-4597 
96495 
3-5988 
3-4728 
96.499 
9.4OIO 
9.0726 
96.5O7 
O.7199 
0.6950 
96.54I 
Mean, 96.512 
Wt. Ag. 
Wt. BaCla. 
Ratio. 
6-59993 
6.36974 
96.512 
5.55229 
5.36OIO 
96.539 
4.06380 
3-92244 
96-522 
Mean, 96.524 
Second Series. 
Third Series. 
Wt. Ag. 
Wt. BaCla. 
Ratio. 
4-4355 
4.2815 
96.528 
2.7440 
2.6488 
96-53I 
6.1865 
5-9712 
96-520 
3-4023 
3-2841 
96-526 
Mean, 96.526 
1 Proc. Amer. Acad.. 29, 55. 206 
F. W. CLARKE. REPORT OF COMMITTEE 
Wt. Ag. 
Wt. BaClj. 
Ratio. 
6.7342 
6.50022 
96.525 
IO.6023 
IO.23365 
96.523 
Mean, 96.524 
Fourth Series. 
All the weights represent vacuum standards. From the four 
series the atomic weight of barium is deduced as follows ; when 
O = i6. 
First series Ba = 137.419 
Second " " 137.445 
Third " " 137-449 
Fourth " " 137.445 
In three more series of experiments Richards determined the 
ratio between 2AgCl and BaCl2. The data are subjoined, with 
the ratio 2AgCl : BaCl2:: 100: x appended. 
First Series. 
Wt. AgCl. 
Wt. BaCla. 
Ratio. 
8.7673 
6.3697 
72.653 
5-1979 
3-7765 
72.654 
4-9342 
3-5846 
72.648 
2.0765 
I.5085 
72.646 
44271 
3-2163 
72-650 
Mean, 72.649 
Second Series. 
Wt. AgCl. 
Wt. BaCl2. 
Ratio. 
2.09750 
I.52384 
72.650 
7.37610 
5.36OIO 
72.669 
5-39906 
3.92244 
72.650 
Mean, 72.6563 
Wt. AgCl. 
Wt. BaCla. 
Ratio. 
8.2189 
5-97123 
72-6524 
4-5199 
3.28410 
72.6587 
Mean, 72.6555 
Third Series. 
Hence we have for Ba, 
First series Ba = 137.428 
Second " " =137.446 
Third " " =137.444 
The mean of all is 137.440, as against 137.434 found in the 
work on the bromide. By combining the two chloride ratios, ON ATOMIC WEIGHTS, PUBLISHED DURING 1894. 
207 
Aga: BaCl2 and 2AgCl: BaCl2, the ratio Ag: Cl can be computed. 
This gives Ag = 107.930, a value identical with that of Stas. 
Cobalt and Nickel.-The atomic weights of these two metals 
have been redetermined by Winkler,1 who adopts a radically 
new method, using the pure electrolytic elements as a starting- 
point. In each case, the weighed metal, deposited upon plati- 
num, is treated with a weighed excess of iodine dissolved in 
potassium iodide. The metals are thus converted into iodides, 
and the excess of iodine is then measured by titration with thio- 
sulphate solution. Thus the direct ratios, Co: I, Ni: I, are 
determined. Two series of estimations are given for each metal, 
with results as follows. The atomic weights used in calculation 
are H - 1, 1= 126.53. 
First Series-Cobadt. 
Wt. Co. 
Wt. I. 
At. Wt. Co. 
0-4999 
2.128837 
59-4242 
0.5084 
2.166750 
59-3772 
O.529O 
2-254335 
59-3828 
0.6822 
2.908399 
59-3582 
0.6715 
2.861617 
59-3824 
Mean, 59.3849 
0.5185 
. 2.209694 
59-3798 
0.5267 
2.246037 
59-3430 
0-5319 
2.268736 
59-3294 
Mean, 59-35°7 
Second Series-Cobadt. 
Mean of all, €0 = 59.3678. 
Wt. Ni. 
Wt. I. 
At. Wt. Ni. 
0.5144 
2.217494 
58.6702 
O.4983 
2.148502 
58.6918 
O.5265 
2.268742 
58.7268 
0.6889 
2.970709 
58.6828 
O.6876 
2.965918 
58.6678 
Mean, 58.6878 
First Series-Nicked. 
Second Series-Nicked. 
O.5I2O 
2.205627 
58.7436 
0.5200 
2.204107 
58.7432 
0.5246 
2.259925 
58.7432 
Mean, 58.7433 
^Ztschr. anorg. Chem.,i, i. 208 
F. W. CLARKE. REPORT OF COMMITTEE 
Mean of all, Ni =58.7155. 
For O = 16, these become 
Co =59.517 
Ni = 58.863. 
Palladium-In 1889 Keiser published his determinations of 
the atomic weight of palladium, for which, since then, other 
investigators have found somewhat different values. He has 
now, jointly with Mary B. Breed, given a new set of determina- 
tions, which confirm his former series.1 As before, palladiam- 
monium chloride was reduced in hydrogen, the salt being pre- 
pared by two methods and carefully examined as to purity. 
Two series of experiments are given, with the following weights 
of material : 
First Series. 
Pd(NH3Cl)2. 
Pd. 
At. Wt. Pd. 
I.60842 
0.80997 
106.271 
2.08295 
1.04920 
106.325 
2.02440 
i-01975 
106.334 
2.54810 
1.28360 
106.342 
I-755O5 
0.88410 
106.341 
From sum 
of weights, 106.325 
Reduced to 
vacuum, 106.246 
Second Series. 
Pd(NH3Cl)2. 
Pd. 
At. Wt. Pd. 
I-5O275 
O.75685 
106.297 
I.23672 
0.62286 
106.296 
1.34470 
O.67739 
IO6.343 
I.49O59 
0.75095 
106.353 
From sum of 
weights, 106.322 
Reduced to vacuum, 106.245 
The atomic weight was computed with H= i, N = 14.01, and 
01=35.37. If O= 16 this becomes Pd= 106.51. This is only 
0.02 less than the value obtained in the earlier investigation. 
Tungsten.-A new determination of the atomic weight of 
tungsten, by Pennington and Smith,2 leads to a much higher 
value than that commonly accepted. The older work seems 
very probably to have been done upon material contaminated 
1 Am. Chern.J., 16, 20. 
2 Read before the Amer. Philos. Soc., Nov. 2, 1894. ON ATOMIC WEIGHTS, PUBLISHED DURING 1894. 
209 
with molybdenum, an impurity which was eliminated in this 
investigation by Debray's method,-that is, by volatilization by 
means of gaseous hydrochloric acid. The metal, carefully puri- 
fied, was oxidized in porcelain crucibles, with all necessary pre- 
cautions, and the following data are given: 
wt. w. 
wt. o3. 
At. Wt. W. 
0.862871 
0.223952 
184.942 
0.650700 
0.168900 
184.923 
0.597654 
0.155143 
184.909 
0.666820 
0.173103 
184.902 
0.428228 
0.111168 
184.9OO 
0.671920 
0.174406 
184.925 
0.590220 
0.153193 
I84.933 
0.568654 
0.147588 
184.943 
1.080973 
0.280600 
184 913 
Mean, 184.921 
All weights are reduced to a vacuum, and O - 16 is taken as 
the standard of reference. 
Another paper, by Smith and Desi, was read at the same 
meeting with that just cited. In this research, the tungstic 
oxide was purified in the same way, and reduced by heating in 
a stream of pure hydrogen. The water formed was weighed, 
and all weights reduced to a vacuum. Computed with 0 = 16 
and H = 1.008, the results are as follows: 
wt. wo3. 
wt. h2o. 
At. Wt. W. 
0.983024 
0.22834 
184.683 
0.998424 
0.23189 
184.709 
1.008074 
0.23409 
184.749 
0.911974 
0.21184 
184.678 
0.997974 
0.23179 
184.704 
1.007024 
0.23389 
184.706 
Mean, 184.704 
Why this result should be lower than that previously found by 
Pennington and Smith remains to be explained. 
Thallium.-Two determinations of atomic weight were made 
by Wells and Penfield to ascertain the constancy of the element 
as such.1 The nitrate was fractionally crystallized until about 
one-twentieth remained in the mother-liquor, while another 
^Am.J. Sci., [3], 47, 466. 210 
F. W. CLARKE. REPORT OF COMMITTEE 
twentieth had been subjected to repeated recrystallization. Both 
fractions were converted into thallium chloride, which was dried 
at ioo°, and in both the chlorine was estimated by weighing as 
silver chloride on a Gooch filter. The results were as follows : 
T1C1. 
AgCl. 
At. Wt. Tl. 
Crystals 
.... 3.9146 
2-3393 
204.47 
Mother-liquor. • ■ • . 
•••• 3.3415 
I.9968 
204.47 
Calculated with Ag = 107.92 and Cl = 35.45. 
In the report for 1893 Lepierre's work on thallium was given, 
and the last value cited was Tl= 203.00, varying widely from 
the rest of the series, and affecting the mean. The mean stated 
by Lepierre was 203.62, and as found by me was 203.57. 
Lepierre1 now calls attention to the fact that his value 203.00 
was a misprint for 203.60, and that his mean was therefore cor- 
rectly given. He also gives additional details relative to his 
work. 
Bismuth.-The long-standing controversy between Schneider 
and Classen over the atomic weight of bismuth, has led to a new 
set of determinations on the part of Schneider.2 The old method 
was still used ; namely, of converting the metal into the trioxide 
by means of nitric acid and subsequent ignition of the nitrate ; 
but the metal itself was carefully purified. Results as follows : 
Wt Bi. 
Wt. Bi2O3. 
Per cent. Bi in Bi2O3. 
5-0092 
5-5868 
89.661 
3-677O 
4.1016 
89.648 
7-2493 
8.0854 
89-659 
9-2479 
10.3142 
89.662 
6.0945 
6-7979 
89-653 
12.1588 
I3-56IO 
89.660 
Mean, 89.657 
If 0 = 16, Bi ranges from 207.94 to 208.15, or in mean 208.05, 
confirming the earlier determinations. 
Tin.-Incidentally to his paper on the white tin sulphide 
Schmidt gives one determination of the atomic weight of the 
metal.3 
0.5243 gram Sn gave 0.6659 SnO2. Hence 811=118.48. 
1 Bull. Soc. Chim., [3], 11, 423. 
2 J. prakt. Chern., [2], 50, 461. 
3 Ber d. chem. Ges., 27, 2713. ON ATOMIC WEIGHTS, PUBLISHED DURING 1894. 
211 
Anomalous Nitrogen.-An important discovery has been made 
by Lord Rayleigh, who finds that nitrogen obtained by purely 
chemical methods is perceptibly lighter than that from atmos- 
pheric air.1 Equal volumes of the gas, variously prepared, 
weighed as follows: 
By passing NO over hot iron 2.30008 
" " N.,O " " " 2.29904 
" " AmNO., " " " 2.29869 
For nitrogen from air he found : 
From air passed over hot iron 2.31003 
" " " through moist FeO2H2 2.31020 
" " " over hot copper 2.31026 
Investigating the cause of this anomaly, with the co-operation 
of Ramsay, Rayleigh came to the astonishing results communi- 
cated a few months later to the British Association. It was 
found, in short, that atmospheric air contains a gas heavier than 
nitrogen, and hitherto unknown. Its density, in a sample as 
pure as could be obtained, was 19.09, and it was characterized by 
extraordinary inertness. Whether it is a new element, or allo- 
tropic nitrogen, Na, remains to be determined. The work is 
cited here because it shows that the density of nitrogen as hith- 
erto determined, can give no trustworthy value for the atomic 
weight of the element. 
Miscellaneous Notes.-Some data bearing upon the atomic 
weight of tellurium are given by Gooch and Howland.2 As the 
homogeneity of tellurium is still uncertain, I omit their details. 
Wanklyn's attempt to show that the atomic weight of carbon 
is not 12, but 6, was noted last year. He has since published 
more on the subject in a paper on Russian Kerosene,3 and the 
matter was also discussed at the Oxford meeting of the British 
Association.4 
In a communication upon the Stasian determinations,0 Hin- 
richs discusses the availability of silver as a secondary standard 
in the scale of atomic weights. He makes silver, chlorine, bro- 
mine, iodine, and sulphur all Proutian in value. Hinrichs also 
1 Chem. News, 69, 231, May 18, 1894. 
lAm.J. Sei., [3], 48, 375. 
KPhil. Man:., [5], 37. 495- 
4 Chem. News, 70, 87, Aug. 24, 1894. 
5 Comfit, rend., 118, 528. 212 
REPORT OF COMMITTEE ON ATOMIC WEIGHTS. 
has published his views upon atomic weights in extenso in book 
form.1 
In conclusion I submit a table of atomic weights revised to 
January i, 1894. O= 16 is still retained as the base of the sys- 
tem ; but I hope that in another year it will be practicable to 
return to H = 1. 
Name. 
Atomic Weight. 
Name. 
Atomic Weight 
Aluminum 
27. 
Neodymium 
••'•••• 140.5 
Antimony 
120. 
Nickel 
58-7 
Arsenic ........ 
75- 
Nitrogen 
14-03 
Barium 
137-43 
Osmium 
190.8 
Bismuth ....... 
208. 
Oxygen 
16. 
Boron 
11. 
Palladium 
106.5 
Bromine ...!... 
79-95 
Phosphorus 
3i- 
Cadmium 
112. 
Platinum 
195- 
Caesium 
132-9 
Potassium 
39-ii 
Calcium 
40. 
Praseodymium •.. 
■■■■■■■ 143-5 
Carbon 
12. 
Rhodium 
103- 
Cerium 
140.2 
Rubidium 
85.5 
Chlorine 
35-45 
Ruthenium 
101.6 
Chromium 
52.1 
Samarium 
150. 
Cobalt 
59-5 
Scandium 
44- 
Columbium .... 
94- 
Selenium 
79- 
Copper 
63.6 
Silicon 
28.4 
Erbium 
166.3 
Silver 
107.92 
Fluorine 
19- 
Sodium 
23.05 
Gadolinium .... 
156.1 
Strontium 
87.66 
Gallium 
69- 
Sulphur 
32.06 
Germanium • • • - 
723 
Tantalum 
182.6 
Glucinum 
9- 
Tellurium 
125. 
Gold 
197-3 
Terbium 
160. 
Hydrogen 
1.008 
Thallium 
204.18 
Indium 
"3-7 
Thorium 
232.6 
Iodine 
126.85 
Thulium 
170.7 
Iridium 
193-i 
Tin •• 
ii9- 
Iron 
56. 
Titanium 
48. 
Lanthanum 
IS8-2 
Tungsten 
184-9 
Lead 
206.95 
Uranium 
239.6 
Lithium 
7.02 
Vanadium 
5i-4 
Magnesium 
24.3 
Ytterbium 
173- 
Manganese 
55- 
Yttrium 
89.1 
Mercury 
200. 
Zinc 
65.3 
Molybdenum ... 
96. 
Zirconium 
90.6 
iThe True Atomic Weight of the Chemical Elements, and the Unity of Matter. By 
Gustavus Detlef Hinrichs. St. Louis. 1894.