ROBERT'S TEST PLATES. 



26. The erroneous estimates of the relative efficiency and power 

 of different microscopic instruments which would result from the 

 use of such test-objects, are obvious. A microscopist in London, 

 observing the tracery of the scale of a Podura, and another at 

 New York observing another scale of the same insect, the former 

 failing to see its striae, which would be visible to the latter, it 

 could not at all be safely inferred, that the instrument of the one 

 was inferior in efficiency and power to that of the other ; and it 

 might even happen, that the instrument which failed to show the 

 striae in London, was, nevertheless, superior to that which ren- 

 dered them distinctly visible in New York. The result of such a 

 comparison would entirely depend upon the structure of the two 

 scales adopted as tests, which might differ within very wide limits. 



Independently of this uncertainty attending the application of 

 such tests, there is another not less serious objection to them ; 

 they hold out a temptation to microscope makers who supply them 

 with the instruments they sell, to select such only as are most 

 easily rendered visible ; and although it be true that this is an 

 expedient to which the most respectable class of makers would not 

 resort, it is nevertheless true that the inferior makers do so, and 

 thereby do injustice to those who are above such practices. 



Natural objects, therefore, do not supply such permanent and 

 unalterable tests for the microscope as the double stars, stellar 

 clusters, and nebulae do for the telescope ; and this circumstance 

 has directed the attention of the higher class of artists, to the pro- 

 duction of artificial test-objects which shall have determinate and 

 certain qualities, and which, like manufactured articles, may be 

 reproduced with such absolute identity as to supply standards of 

 comparison that can be applied in different places, and at different 

 times, to different instruments, so as to give results which will 

 admit of comparison. 



27. The production of micrometer scales, by Mr. Froment, the 

 divisions of which are separated by intervals so small as the 

 25000th of an inch, has been already mentioned. 



Now the lines marking such divisions being in closer proximity 

 than those of the tracings upon certain test-objects, it will be 

 evident that artificial test- objects might be made by means similar 

 to those by which such scales have been executed, and there can 

 be little doubt that the great artistic skill which has succeeded in 

 producing traces, separated by the small interval above named, 

 could be pushed further, so as to produce striated surfaces, which 

 would serve all the purposes of test-objects. 



Mr. Nobert, of Griefswall, in Prussia, has taken up this problem 

 of test-objects, and, without attempting, as it would appear, to 

 engrave micrometric scales, which would require intervals of some 



F 2 67 



