NOBERT S TEST-PLATE. 



139 



Amphipleura, as stated by Kabenhorst, transverse stria? are at the 

 rate of 135 in -001" (according to which the distance of the 

 markings amounts to only *187 mic.) is evidently a mistake. 



Nobert's Test-plate. 



The employment of organic test-objects is always attended by 

 the disadvantage, that the results which the observer obtains are 

 never exactly comparable with those of another observer, objects 

 of the same kind differing so much from one another in size and 

 distinctness. To overcome this inconvenience, Nobert invented a 

 method of producing an artificial test-object, in the form of a 

 micrometric series of lines upon glass. In 1846 he ruled test- 

 plates with ten bands of lines, in which the distances of the lines 

 were greatest in the first band, and progressively less in each 

 successive band. Subsequently he increased the number of bands 

 to twelve, fifteen, twenty, and, later on, even up to thirty, at the 

 same time altering the distances from those in the older plates, so 

 that the results obtained with the latter cannot be accurately 

 compared with the results obtained with the modern plates. 

 Nobert has recently reduced the number of the bands to 

 nineteen. The distances of the lines in the two most recent 1 test- 

 plates, and their number in a millimetre, are collected in the 

 following tables : 



A. THIRTY-BAND PLATE. 



1 Still more recently Xobert has produced a test -plate on which are ruled 

 twenty bands of lines ; those lines ruled in the tenth band are equivalent in 

 fineness of division to the nineteenth group described in Table S ; and those in 

 the twentieth group are of twice that fineness of division. [Er.] 



