THE MICROSCOPE. 



minute upon the glass, with the necessary precision, especially 

 when it is remembered that any error or inequality will necessarily 

 be augmented in the exact proportion of the magnifying power 

 with which such a scale is seen. Nevertheless this difficulty has 

 been most successfully overcome, and combinations of screws and 



Fig. 24. 



wheels have been contrived, by which the diamond point is moved 

 by self-acting mechanism, so as to trace the successive divisions 

 of scales of astonishing minuteness. Scales are thus produced, 

 the divisions of which are no greater than the 25000th part of 

 an inch. 



This extreme minuteness is, however, rarely necessary or desira- 

 ble in microscopic researches, and the divisions of the scales in 

 more common use vary from the 1000th to the 2000th of an inch. 

 In the scales delivered with moderately good French instruments, a 

 millimetre is divided into one hundred parts. A millimetre being 

 about the 25th of an inch, these divisions would therefore be the 

 2500th of an inch. (See Tract on Microscopic Drawing and En- 

 graving, Museum, vol. vi.) 



The process described above, in which the object is measured 

 by superposition upon the micrometric scale, is attended with 

 several practical difficulties and objections. The object, when 

 thus placed, is always nearer to the object-glass than the scale, 

 and when it is in focus, the scale is out of focus and invisible ; 

 and, on the other hand, when the scale is in focus, the object is 

 out of focus and indistinct. "When low powers only are used, 

 this difference between the focus of the object and that of the 

 scale being inconsiderable, will not prevent the success of the 

 operation ; but when the powers are high, it can never be satis- 

 factorily, and sometimes not at all effected. 



There is still another objection to the process. The placing and 

 displacing of objects frequently on a surface so delicatety engraved, 

 subjects it to friction, which soon spoils and effaces the divisions. 



If the divided surface be protected, as it may be, by a plate of 

 glass laid upon it, the difference between the distances of the 

 object and the scale from the object-glass is augmented by the 

 52 



