and that there is certain structure which is not 

 visible, how can we reconcile this with the fact 

 that numerical aperature only does it? Why, if 

 this is true, should we use a higher power object- 

 ive with a certain aperture in preference to a lower 

 one of the same aperture? 



The normal eye can see about 200 lines to the 

 inch, or structures which are ^-^th of an inch 

 apart and it is therefore evident that any structures 

 under the microscope must be separated at least 

 to this extent in the virtual image, in order that 

 they may become visible. While we have shown 

 that magnifying power may be increased 

 by increasing power of objective, 

 by increasing power of eyepiece-, 

 by increasing tube length, 



we are limited in the last direction mainly by 

 such length of tube as has been found conveni- 

 ent in the construction of the stand; in the case of 

 eyepiece on the one hand by convenience in use 

 and on the other by the fact that too great an increase 

 causes an indistinctness, so that although variation 

 in eyepiece within narrow limits is admissible, we 

 are compelled to select an objective which, with a 

 medium power eyepiece, will give such separation 

 to structure and fine detail, that it will be visible 

 to the eye without any undue strain. 



