'CRITICAL' AND UNCRITICAL IMAGES 421 



objective with a dark-ground illumination, employing a binocular, is 

 very suitable. Stained bacteria in tissue are best seen with a l<i !</< 

 cone, as was pointed out by Dr. Robert Koch, and is directly supported 

 by Dr. Abbe as suitable in his directions for the use of the Abbe 

 condenser. 1 The brilliancy of the illumination obliterated the thin 

 tissue which is in a medium whose refractive index is similar to 

 itself. The bacteria, which are opaque with pigment, then stand 

 out boldly. A bacterium not in tissue is always better seen by 

 means of a large cone, provided that the objective is properly 

 corrected. The very minute hairs on the lining membrane of the 

 blow-fly's tongue, if examined by a ^ objective and a narrow cone, 

 appear thickened, shorter, more blunted, and often split into two 

 parts. This is shown in figs. 1 and 3 in the frontispiece. Fig. .'! is 

 a critical image magnified 510 diameters. A lens should be used to 

 examine this. 



It will be seen that the hairs, especially the long central one. are 

 very fine and spinous. They have not the ring socket common to 

 insect hairs, but grow directly from a delicate membrane. 



This photograph was taken with an apochromatic J of - 95 X.A. 

 and No. 3 projection eye-piece ; and it was illuminated by means of 

 a large solid cone of '65 N.A. from an achromatic condenser. 



Fig. 2 is an uncritical image, with all the conditions as above. 

 s;ive that a cone of small angle, i.e. of O'l, was used for illumination. 



The first alteration which thrusts itself upon the eye is the 

 'foiifilitti/ of the hairs which are in the least degree out of focus. 

 Hut, further, it will be noted that there is a bright line with a dark 

 edge round the hairs which are precisely in focus ; this is a diffraction 

 effect, always, in our experience, present in objects illuminated by 

 cones of insufficient angle, and it can be easily made to disappear by 

 widening the cone. As the illuminating cone is enlarged they become 

 sharper and longer, and their edges become more definite. But 

 nothing is gained, but rather a distinct loss is incurred, by making 

 the illuminating cone much larger than three-fourths of the objective 

 cone. 



As an example of erroneous interpretation, the representation of 

 the pygidium of a flea by some leading sources of information of a 

 few years ago may be instanced. It was a special test of many 

 authors, and has been carefully figured ; this shows that it is not an 

 accidental error, which it might have been if it were merely an 

 ordinary object ; it is an error depending in all probability on a 

 faulty system of illumination. Moreover, the error cannot be attri- 

 buted to the object-glasses of the time, as it is a low-power object, 

 and the low powers of that day were quite as good as those lately in 

 use. In the descriptions and in the drawings, often beautifully 

 executed, the hairs proceeding from the centre of the wheel-like 

 discs are represented as being ' stiff and longish bristles, thick at 

 one end and tapering off to a point. And the small hairs round 

 are described as ' minute spines ; ' in the drawing they are like the 

 spinous hairs of an insect, and have the usual socket-joint at the 



1 Directions for the Use of Abbe's Illuminating Apparatus a leaflet issued by 

 Carl Zeiss, 1888. 



