2.26 WOODWARD, ON NOBERt's TEST-PLATE. 



According to Mr. Beck, the lines of the 20th band are 

 thirty-five in number, and are ruled at the rate of 70,000 to 

 the English inch, which corresponds almost precisely with 

 the statement of Starting. 



Nobcrt subsequently prepared a test-plate with thirty 

 bands, the lines of the 1st being the -j-f^th, those of the 

 30th the -sTy'TTTith of a Paris line apart. He states that the 

 lines are ruled at the following rates, for the bands named : 



No. of lines to No. of lines to 



a niilliiriefer. a millimeter. 



No. 1 443 No. 20 2653 



5 806 25 3098 



10 161,2 30 .... 3544 



15 2215 



The 20th band of the twenty-band plate corresponds 

 nearly with the 22nd band of this plate. 



An analysis of this thirty-band plate has been made by 

 Messrs. Sullivant and Wormley,* who succeeded satisfactorily 

 in resolving the first twenty-seven bands, and counting the 

 lines in them. Up to the 26th band they encountered 

 " no serious difficulty in resolving and ascertaining the posi- 

 tion of the lines ; but on this and the subsequent ones 

 spectral lines, that is, lines each composed of two or more 

 real lines, more or less prevailed, showing that the resolving 

 power of the objective was approaching its limit. By a suit- 

 able arrangement, however, these spurious lines Avere sepa- 

 rated into the ultimate ones on the whole of the 26th, and 

 very nearly on the whole of the 27th band ; but on the 28th, 

 and still more on the 29th, they so prevailed that at no one 

 focal adjustment could more than a portion (a third or a fifth 

 part) of the width of these bands be resolved into the true 

 lines. The true lines of the 30th band we were unable to 

 see, at least with any degree of certainty." 



Still more recently Nobcrt has prepared the plate of nine- 

 teen bands, mentioned at the commencement of this article. 



The following statement of the distance of the lines in the 

 several bands of this plate, with the number of lines to the 

 millimeter for each, is taken from Starting.f 



No. of band. Distance of lines. !J°- ^^ )!>ies to 



the millimeter. 



1 TTiVo of a Paris line 443 



2 Wk^ „ 665 







* " Oh Nobert's Test-plates, &c.," by W. S. Sullivant and T. G. Worm 

 ley. 'American Journal of Science and Arts' for January, ISOl. 

 f Loc. cit., p. 374. 



