Intelligence and Miscellaneous Articles. 571 



line H (3. The bright line of the gas was projected outside the dark 

 line of the star, over its border on the side towards the red : the 

 amount of this superposition could not be estimated exactly ; for 

 the stellar line (or, rather, band) is very broad, and stumped at its 

 edge, as we have ascertained from the commencement of our re- 

 searches. We are surprised that observers do not dwell upon this 

 aspect of most of the stellar hydrogen-lines. We shall be pardoned 

 for noting all these details, considering the importance of the con- 

 sequences which now from them. 



This result was, in the main, the same as that which we had 

 already obtained, especially in 1869, and published in our memoirs 

 Sur les Spectres Solaires in 1872, p. 44. The star seemed, then, to 

 have its undulations shortened, and, consequently, to be approaching 

 us, as they found at Greenwich, contrarily to Mr. Huggins. 



Not trusting my own estimation, I had the observation made by 

 my assistants, all of them accustomed to these measurements. Ob- 

 servation always gave the same result when the telesoope was carried 

 by the clock, and when moreover an assistant was at the searcher 

 to keep it upon a fixed point corresponding to the slit of the spec- 

 troscope ; but if the clock stopped, or if the assistant deranged the 

 position of the star, the bright line was seen accidentally to place 

 itself on the other side or in perfect coincidence with the dark line. 

 This induced us to repeat more carefully the observation the follow- 

 ing day, in order to discover the cause of these variations. 



In this second series the results were at first similar to those of 

 the preceding day ; but on abandoning the use of the clock the line 

 appeared to take its place on one side or the other, according as the 

 star was struck or carried away on one side or the other with respect 

 to the axis of the telescope. The clock was then again applied, 

 when the bright line appeared on the side towards the red ; and 

 after the spectroscope was turned on its axis 180°, the line was 

 situated on the other side for certain observers, remaining on the 

 former side for others. 



After producing again and again for a long time these different 

 changes of position, we acquired the conviction that the line miglit 

 appear constantly on one side or the other, according to the position of 

 the instrument, without the observer having any sufficiently certain 

 indication to cause him to recognize the illusion of which he ivas the 

 victim. This is plainly a very grave conclusion. We do not pre- 

 tend that observers so skilful have been deceived ; but we point out 

 a possible cause of error in this class of observations. We have not 

 yet succeeded in discovering the source of these changes, which may, 

 perhaps, be attributable to a species of parallax, owing to the focus 

 of the star-image not coinciding with the plane of the slit. We 

 confine ourselves to merely indicating the phenomenon which might 

 well have produced an illusion for others as it deceived us. Here 

 the concordance of frequently repeated observations would give no 

 assistance ; for, in looking, one always places one's self so as to have 

 the clearest view — that is to say, in such a position that the same 

 error is reproduced. — Oomptes Rendus de VAcademie des Sciences, 

 April 3, 1876, vol. lxxxii. pp. 761-765. 



