JResolving-power of Telescopes. 117 



a telescope of 10 centimetres aperture on a distant scale illu- 

 minated by direct sunshine. In this form the experiment is 

 troublesome and requires expensive apparatus — difficulties 

 which are entirely obviated by the plan which I have followed 

 of using a much smaller aperture. 



The object, on which the resolving-power of the telescope is 

 tested is a grating of fine wires, constructed on the plan em- 

 ployed by Fraunhofer for diffraction-gratings. A stout brass 

 wire or rod is bent into a horseshoe, and its ends are screwed. 

 On these screws fine wire is wound of diameter equal to about 

 half the pitch, and secured with solder. The wires on one 

 side being now cut away, we obtain a grating of considerable 

 accuracy. A wire grating thus formed is preferable to a scale 

 ruled on paper, and placed in front of a lamp presents a very 

 suitable subject for examination. The one that I employed 

 has 50 wires to the inch, and for security is mounted in a 

 frame between two plates of glass. For rough purposes a 

 piece of common gauze with 30 or 40 meshes to the inch may 

 be substituted with good effect. 



For the sake of definiteness of wave-length the grating was 

 backed by a soda-flame, though fair results are obtainable 

 with a common paraffine-lamp. The telescope is a small in- 

 strument mounted on a stand, and provided with a cap by 

 means of which various diaphragms can be conveniently fitted 

 in front of the object-glass. The apertures in these diaphragms 

 may be either circular or rectangular. In the latter case the 

 length of the slit is placed parallel to the wires of the grating, 

 and we have the advantage of greater illumination than with 

 a circle of equal width. The observation consists in ascer- 

 taining the greatest distance at which the wires can be seen 

 resolved. For this purpose the telescope, focused all the 

 while, is gradually drawn back until in the judgment of the 

 observer the periodic structure is no longer seen;, and the dis- 

 tance between the grating and the diaphragm is then measured 

 with a steel tape. The distance thus determined is more de- 

 finite than might be expected, the differences in the case of 

 various observers not usually amounting to more than 2 or 3 

 per cent. 



Two slits were tried, half an inch long, and of widths '107, 

 •196 inch respectively. These widths were measured by in- 

 serting a graduated wedge. It was found, however, that the 

 graduations could not be trusted; so that the wedge was in 

 fact used merely to convey the length to be measured to a pair 

 of callipers reading to one thousandth of an inch. The dis- 

 tances at which resolution just ceased were estimated respec- 

 tively as 91*5 and 168'5 inches, corresponding to angular 



