RELATIVE POWER OF DISCRIMINATION. 147 



diameter of the sensitive retinal elements; 1 for it cannot be 

 supposed that these latter render qualitatively different excitations 

 appreciable. In the last instance, therefore, the visibility of a 

 dark line upon a bright ground depends chiefly upon whether 

 the eye of the observer is still capable of appreciating a slight 

 difference of light, which is dispersed over a part of the retinal 

 elements. The supposed testing of the Microscope virtually passes 

 into a testing of the eye. 



Under these circumstances we refrain from giving a table of 

 our measurements, and refer those who take a special interest 

 in such observations to the detailed work of Harting. 



The testing of objectives by means of diatom- valves, or Robert's 

 test-plates, as regards the relative power of discrimination, leads, 

 naturally, to the same result, cceteris paribus, as the comparison of 

 the images of wire-gauze meshes that are still just perceptible, 

 always on the supposition that, to a certain degree, the different 

 series of lines admit of comparison. The limit of discrimination 

 usually appears to be slightly displaced, owing to the varying 

 nature of the details and the unequal method of illumination. 

 Thus, for instance, it is a fact that diatoms with faint markings, 

 although their distance from each other is exactly the same, 

 are always more difficult to resolve than those with strongly 

 marked series of striae. Amplupleura pellucida is for this reason 

 more difficult than Frustulia, saxonica, and Navicula rliomboides 

 than Nitzschia sigmoidea, &c., although the distances of their 

 markings are approximately equal. This is also the reason 

 why organic test-objects are much less adapted for comparative 

 examinations than images of wire-gauze, or the series of lines 

 on Nobert's plates. 



1 The "bacillary layer," which is regarded as the true light-perceiving 

 organ, consists in the yellow spot (the point of clearest vision) merely of so- 

 called " cones," whose diameter amounts to about 5 mic. This magnitude, if 

 the distance of the focal point of the retina is taken at 15 mm., corresponds to 

 a vision of about 68", and with an object distance of 25 cm. to a diameter of 

 83 mic. According to Harting (" Mikroskop," p. 298 ; 2nd ed. i. p. 333), 

 thread-shaped objects, or rather their images formed by air-bubbles, are 

 microscopically perceptible even with an amplification of 50, if their actual 

 thickness equals -194 mic., and consequently the diameter of the microscopic 

 image equals 50 X '194 = 9 '7 mic. The corresponding retinal image would 

 not, therefore, occupy even the eighth part of one retinal element. 



