IMAGES OF WIRE-GAUZE AS TEST-OBJECTS. 127 



or only indistinctly. We will not dispute that the data thus 

 obtained are perfectly sufficient, on the whole, for a correct esti- 

 mation of the capacity of a given instrument. We must not, 

 however, overlook the fact that these test-objects themselves may 

 easily lead to an entirely one-sided conception of the optical 

 power, because the pencils of light, by whose interference the 

 details are delineated, with the ordinary illumination pass through 

 only a narrow peripheral zone of the objective, and consequently 

 disclose nothing as to the whole central portion of the refracting 

 surfaces. We learn therefore, as a rule, merely the limits of 

 resolution, not the optical power in general. This also applies to 

 the images of wire-gauze, produced by small air-bubbles, &c., if 

 the meshes are diminished to the limit of discrimination. The 

 microscopic image is here also an interference image which, as 

 regards its details, is formed exclusively by the pencils of light 

 which graze the margin of the ap^ture of the objective. Never- 

 theless, these images of wire-gauze (which Harting has recom- 

 mended) offer certain advantages which the ordinary test-objects 

 do not afford, and to which we are inclined to assign special value, 

 especially for comparative observations, and hence we place them 

 in the first rank in the following considerations. 



Images of Wire-gauze as Test-objects. 



The images of wire-gauze are recommended as test-objects prefer- 

 ably to all others because, in the first place, for every range of 

 amplification they present one and the same object ; for the virtual 

 images of a spherical air-bubble, always formed with equal sharp- 

 ness, can be diminished beyond the limit of discrimination, without 

 alteration of the relations of light and shadow. The pencils of 

 light proceeding from the wire-gauze which, on account of the 

 minuteness of the air-bubble, are to be regarded as composed of 

 parallel rays under all circumstances undergo (as shown in Fig. 

 80) such a refraction that, whilst completely filling a widely 

 opened cone, they diverge towards the objective. (The Fig. is 

 accurately constructed for refraction in water at the calculated 

 angles.) 



To this agreement with test-objects which so far as regards the 



