46 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 54 



1807 series. . 190Jf series. 



3.6359 3.63G8 

 3.6354 3.6360 

 3.6364 3.6362 

 3.6358 3.6363 



3.6360 3.6367 

 3.6366 



Mean, 3.6359 3.6354 



Mean, 3.6363 



These are so nearly together that I venture to treat them as one 

 series, in mean 3. 6361, ±.000093. The weight of the same volume of 

 oxygen was 2. 6276, ±.00004. The value of the ratio, therefore, is 32: 

 44.2819, ±.0037. 



The measurements by Guye and Pintza ^ are stated so as to show the 

 weight of a normal litre of nitrous oxide. The figures are, in grammes — 



1.97762 

 1.97707 

 1.97760 



Mean, 1.97743, ± .00015 



The weight of a litre of oxygen, according to Morley, is 1.42896, ± 

 .000028. Combining this with Guye and Pintza's figure the ratio becomes 

 32: 44,2824, ±.0035. 



The three independent values for the density ratio Og : N,0, combine 

 as follows : 



Leduc 44.3050, ± .0037 



Rayleigh 44.2819, ± .0037 



Guye and Pintza 44.2824, ± .0035 



General mean 44.2895, ± .0021 



This mean corresponds to a normal litre-weight for nitrous oxide of 

 1.97775 grammes. 



It is convenient at this point to consider the volumetric analysis of 

 nitrous oxide made by Jaquerod and Bogdan.^ A measured volume of the 

 gas was decomposed by an electrically heated spiral of iron wire, and the 

 volume of the residual nitrogen was measured afterwards. Then, with 

 the known densities of the two gases, the ratio between them was easily 

 calculable. Eeduced to uniform conditions, one litre of nitrous oxide 

 gave the following volumes of nitrogen : 



* Compt. Rend., 139, 677. 1904. Corrected in C. R., 141, 51. 1905. The corrected figures are 

 used here. 

 *Journ. Chim. Phys., 3, 562. 1905. 



