ON PHYSIOLOGICAL LIMITS OF MICKOSCOPIC VISION. 467 



How important such an arrangement is for microscope vision may 

 be gathered from the following calculations, it being admitted that 

 the conversion of luminous waves into nerve movement, which con- 

 stitutes the first step in the act of seeing, really does take place in 

 these elements : — 



Let us suppose an object to contain lines or particles distant from 

 each other by the to^Vct o P^^^ ^^ ^^ i^^^> ^^^ ^^^^ ^^ desire to 

 know the requisite magnifying power which shall bring the retinal 

 images of these lines up to such dimensions as would enable them 

 to be distinctly formed on the retina so that the lines or particles 

 shall be perceived separately by the cones. Now an amplification 

 of UisiS) should increase the distance between the lines so that in 

 the microscope image they would be situate ^ Jq of an inch apart, 

 and their retinal image would then shew them -j^Vo ^^^^ apart. 

 This would nearly cover four cones, having a distance of ^-jo o ^^^^ 

 apart, i,e., the amplification of lo^^o would be unnecessarily high 

 for a retinal image thrown on the axial yellow spot. But an amplifi- 

 cation of ^^ would, by a like calculation, give a retinal image in 

 which the particles would stand 3-4^00 ^^^^ apart, and as this corres- 

 ponds with the actual distance between the centres of the finest 

 cones (g-^), the amplification necessary to see lines distant from 

 each other by ^^^^^ inch is only - f^ times. This proves conclusively 

 that the difficulty of resolving lines at 100000 to the inch depends upon 

 other circwnstances than magnifying power ^ and points to the defect 

 of illumination and defining power. 



But, further, the retinal image of an object subtending a visual 

 angle of 60 seconds is 0.00438 m,m. (Bonders), which, converted 

 into a fraction of an inch, is 3- -^-^-q ; and this again corresponds 

 nearly with the limit of distinguishing power of the finest retinal 

 cones, as above given, 3-jVo i^^^- ^^^^ ^^ ^^ ^^1 ^^^ retina does not 

 distinctly appreciate any closer interval than that given hy an object 

 subtending a visual angle 0/ 60 seconds or a few seconds less. 



The question naturally follows — if a higher magnifying power 

 be used does it confuse retinal perception by spreading the retinal 

 image of a line a point over several cones ? The answer is not if 



