18S1.] 



601 



[Chase. 



Scliellen's table gives 393.28 for tbe wave-length of H2. His other valites 

 are precisely the same as the harmonic lengths which are given here. 

 The greatest interval between the successive coincidences of phase in the 

 propagation of the waves which represent any two adjacent lines, is less 

 than -^g of an inch. 



Some of the closer lines precisely represent simple geometric progres- 

 sions within the limits of uncertainty of observation. The middle D line, 

 according to Gibbs, is a geometrical mean between the extreme lines. 



Gibbs, in Jolmsons' Cyclopedia, gives four b lines, the additional line being 

 also geometrically determined : 



Gibbs. Geometric. Log. 



b, 518.31 ar" 518.31 2.7145896 



J., 517.33 ar 517.83 3.7136776 



63 516.85 cw3 516.86 2.7133736 



b^ 516.69 a 516.68 2.7133216 



151. Cotjjoined Harmonies. 

 The indications of geometrical progression in the D and b groups sug- 

 gest the propriety of looking for similar evidence among the remaining 

 lines. The relations which I have pointed out between sethereal and 

 planetary nodes (Note 37), led me to IooIj to the inertia of Sun and 

 Jupiter as an important source of nodal influence upon fetliereal waves. 

 Light-waves would traverse a trajectory equivalent to that of Earth's daily 

 synodic rotation 5.303798 X 365.25636 = 1900.355 times, during the inter- 

 val which would be required for them to traverse one equivalent to Ju- 



