contributions to the Spectral Analysis of the Stars. 361 



spectrum-analysis, have been made in the optical instruments for 

 observing the spectrum, open out the prospect of demonstrating 

 that influence on the spectra of the stars. Theory requires that 

 this should consist of a small displacement of the spectrum-lines, 

 which, for instance, for the mean velocity of the earth of 18*2 

 miles in a second, amounts to the tenth part of the distance 

 between the two sodium-lines. This magnitude, which is very 

 easily deduced from the velocity of light and the length of oscilla- 

 tion of the rays corresponding to the sodium- lines, has been 

 quite recently again deduced by J. C. Maxwell in accordance 

 with earlier calculations by F. Eisenlohr*. 



Yet the magnitude of the displacement appears to Maxwell to 

 be so small, that he concludes his observations with reference to 

 the spectroscopes hitherto constructed and the method of deter- 

 mining the position of the lines with the remark, "it cannot be 

 determined by spectroscopic observations with our present instru- 

 ments, and it need not be considered in the discussion of our 

 observations "f. 



Huggins, nevertheless, in his most recent paper J, of which the 

 above-mentioned investigations of Maxwell form an integrant 

 part, has attempted the solution of the problem in question by 

 using a spectroscope with not less than five prisms, of which 

 two are flint-glass Amicus, and three crown-glass. 



The great enfeeblement of light produced by so great a num- 

 ber of prisms permits the observation of only the brightest stars. 

 Huggins even restricts himself to the communication of his 

 results from observations on Sirius, and thought he had here 

 found a slight displacement of the line F compared with a bright 

 hydrogen-line produced by a Geissler's tube. The direction and 

 magnitude of the displacement would indicate an increase of the 

 distance between the earth and Sirius with a velocity of 41° 1 

 English miles in a second. 



Eliminating the component of the earth's motion, which at the 

 time of observation amounted to twelve miles, the velocity with 

 which Sun and Sirius move apart would be 29*4 miles in a 

 second. 



Huggins himself considers this result as affected with great 

 uncertainty — an uncertainty partly due to the enfeeblement of 

 the light produced by numerous prisms, partly to the difficulty 

 of comparing the coincidences of the bright lines of terrestrial 

 luminous sources with the analogous dark ones of the star- 

 spectra. The latter have at times a different appearance — are, 

 for instance, indistinct at the edges and of variable breadth, as is 

 just the case with this line F in the spectrum of Sirius. 



* Heidelberger Verh. d. phys.med. Ges. vol. iii. p. 190. 

 t Phil. Trans. 1868, p. 532. + Ibid. p. 535. 



Phil. Mag. S. 4. Vol. 38. No. 256. Nov. 1869. 2 B 



