114- SULLTVANT AND •V\'0IIMLEY^ ON NOBERT's TEST-PLATE. 



We add the 3d and 6tli cohimns, giving the distances iu 

 parts of the English inch found by mnltiplying the decimals 

 in the 2d and oth cokimns by -088815. 



Analysis of NoherVs Test-plate of Thirty Bands. 



The figures iu the 3d and 6th columns, showing the dis- 

 tance apart of the lines iu each band, are the mean of numer- 

 ous and slightly variant trials, particularly on the higher 

 bands. Up to the 26th band there was no serious difficulty 

 in resolving and ascertaining the position of the lines ; but 

 on tliis and the subsequent ones, spectral lines,^* that is, lines 

 each composed of two or more real lines, more or less pre- 

 vailed, showing that the resolving power of the objective was 

 approaching its limit. By a suitable arrangement, however, 

 of the illumination, these spurious lines were separated into 

 the ultimate ones on the Avhole of the 26th, and very nearly on 



* The tendency of lines near the limits of the objective's resolving 

 power to run into each other, and produce spectral or spurious lines, is 

 readily shown by a low objective on the lower bands. Hence the mere 

 exhibition of lines is not always conclusive evidence of their ultimate reso- 

 lution. A practised eye will generally distinguish the false from the true. 

 Recourse to a higher objective often accomplishes the same ; but when 

 these fail, t lie micrometer only, together with a previous knowledge of the 

 actual position of the true lines, can determine whether the lines exhibited 

 are real or spurious. A 1-12 orl-lG will show the 3 or 4 highest bands on 

 this plate regularly and beautifully striped with lines much coarser than the 

 true ones ; the same M"ith the 1-30 on the last band. 



