1889.] chief Fluting seen in the Spectrum of Manganese. 37 



The distances of these from the most refrangible edge of Mn (1) 

 are respectively 0004*5 and 0006*0. 



If we take the maximum divergence from the Mn fluting, viz., 

 0006, it is equal to the distance between the D lines of sodium, which 

 are also shown in the diagram for purposes of comparison. The 

 divergence of the other extreme, viz., 5571*5, is about two-thirds of 

 the distance between the D lines. 



Putting Dr. Huggins's observations aside, as having been made 

 with considerable dispersion, any observer who is at all acquainted 

 with the small dispersion generally necessarily employed in observa- 

 tions of the aurora, will quite understand that it is almost impossible 

 to differentiate to these small amounts with such dispersion. The 

 aurora spectrum being usually faint, it is necessary to employ a wide 

 slit, so that the chances of seeing D double are very small. 



Dr. Huggins states that in his observations " the reading showed 

 the line to fall about midway between two strong lines in the spec- 

 trum of tin, X 5564 and 5587 respectively." The wave-lengths given 

 by Thalen for these lines of tin are 5565 and 5588, so that a line 

 exactly half-way between them would have a wave-length of very 

 nearly 5575*5 (whether we take the wave-lengths given by Huggins 

 or by Thalen), which almost absolutely corresponds to the Mn fluting. 

 The value deduced, however, was 5571, which would place the line 

 not half-way between the tin lines, but less than one-third of the 

 distance between them from the most refrangible one. If the state- 

 ment that the line fell midway between the tin lines is to be relied on, 

 it seems probable that there was an accidental shift before the obser- 

 vations of the tin spectrum were made, but the details of the compari- 

 son not having been recorded, it is impossible to say with certainty 

 whether there was any such shift or not. Such a shift would also 

 affect the results of the tellurium comparison referred to by Dr. 

 Huggins to the same extent as that of the tin comparison, since the 

 aurora line was not again brought on the cross- wires. The value 

 determined by Dr. Huggins from a comparison with the iron lines in 

 the neighbourhood was 5571*5, which is a little higher than that 

 derived from the tin and tellurium comparisons. 



[Note, May 7th. — I have received a note from Dr. Huggins referring 

 to the above paragraph, in which I make a quotation from his paper, 

 and suggest a possible source of error. I have his permission to 

 publish the following extracts from his letter : — 



" I thought it would be clear that the statement that the reading 

 showed the line to fall between the lines of tin (or 'about midway 

 between '), was nothing more than a rough indication to me what 

 metal to take, and that no importance was attached by me to' this quite 

 rough determination of the metal to be chosen. 



