ATOMIC WEIGHTS 29 



The details of Noyes' investigation are too voluminous for repetition 

 here. It goes almost without saying that every precaution was taken 

 which his own previous experience and the experience of others could 

 suggest, and that his materials were of the highest degree of purity. The 

 true value of the ratio must lie somewhere within the range of variation 

 shown by his individual determinations, which it may be observed, overlap 

 those of Morley. 



We have now before us, for combination, fifteen sets of determinations 

 of the hydrogen-oxygen ratio. I have arranged them in the order of 

 descending magnitude, and computed their general mean as follows : 



Ratio. Atomic weight H. 



1. Erdmann and Marchand.. 15.975, ±.0113 1.00156 



2. Dumas 15.9607, ± .0070 1.00246 



3. Keiser, 1888 15.9514, ± .0011 1.00305 



4. Thomsen, 1870 15.91, ± .0113 1.00565 



5. Noyes, 1890, uncorrected.. 15.8966, ± .0017 1.00650 



6. Dulong and Berzelius 15.894, ±.0570 1.00667 



7. Rayleigh 15.89, ± .OOyO 1.00692 



8. Leduc 15.881, ± .0132 1.00750 



9. Keiser, 1898 15.8799, ± .0046 1.00756 



10. Morley 15.8790, ± .00028 1.00762 



11. Noyes, 1907 15.8745, ± .00021 1.00783 



12. Cooke and Richards 15.8690, ± .0020 1.00825 



13. Thomsen, 1895 15.8690, ± ,0022 1.00825 



14. Dittmar and Henderson.. 15.8677, ± .0046 1.00834 



15. Keiser, 1887 15.864, ± .0150 1.00857 



General mean 15.8779, ± .00016 1.00769, ± .00001 



In this combination, which includes all the syntheses, good or bad, 

 the general mean lies between the values found by Noyes and Morley. 

 It is, therefore, not far from the truth. If we reject the high values, 

 Nos. 1 to 7, the general mean becomes 15. 8760, ±.00017, and H = 1.00781, 

 ±.00001. Values 10 and 11, combined, give 15. 8761, ±.00017, and H = 

 1.00780, ±.00001. That is, the Morley and Noyes determinations control 

 all the others, and practically eliminate them. The high and low figures 

 tend to balance one another, and so to disappear from the final combi- 

 nation. 



In discussing the relative densities of oxygen and hydrogen gases we 

 need consider only the more modern determinations, beginning with 

 those of Dumas and Boussingault. As the older work has some his- 

 torical value, I may in passing just cite its results. For the density of 

 3 



