ESTIMATING THE EDGES OF MINUTE OBJECTS 277 



The arbitrary scale should be capable of movement by a screw, 

 otherwise the appliance is hardly as accurate as the first method of 

 micrometry by simple drawing described above. 



Of all the methods of micrometry the most accurate is that 

 performed by photo-micrography. A negative of the object to be 

 measured is taken, and then, without any alteration in tube- or 

 camera-length, the magnified image of the stage micrometer is pro- 

 jected on the ground glass ; this is spanned by means of a pair of 

 spring dividers. The negative film is then scratched by these 

 dividers. Then you are in a position to make the most accurate 

 measurement the microscope is capable of yielding. 



It is exceedingly important, when performing micrometric 

 measurements, to remember that the precise edges of all objects in 

 the microscope are never seen. Consequently it is impossible to 

 ascertain from what point to wh^it point the measurement is to be 

 made. 



This, while hardly affecting large and coarse objects, becomes 

 supremely important with small objects. 



Instead of a real edge to an object you get diffraction bands. 

 These bands alter with focus, and also to a greater extent with the 

 angle of the illuminating cone as well as with the aperture of the 

 objective. Hence it ensues that the accurate micrometry of delicate 

 objects presents one of the most difficult matters encountered in 

 practical microscopy. At the present time opinions differ greatly 

 as to the treatment of particular cases. 



The following plan of Mr. Nelson's is the outcome of a long series 

 of experiments : 



1. The focus and adjustment to be chosen may be termed that of 

 the * black dot ' (see Elimination of Errors of Interpretation) ; 

 in other words, if the object were a slender filament it would be 

 represented white with black edges. These black edges are due to 

 diffraction. If the filament is very slender and the illuminating cone 

 small, there may be seen a white diffraction edge outside the black 

 one, and perhaps another faint black one outside that again. 



2. Reduce as far as possible the extent of these diffraction bands 

 by (a) using an objective with as large an aperture as possible ; (b) 

 by using as large an illuminating cone as possible. 



3. Measure from the inner edge of the inner diffraction band to 

 the inner edge of the inner diffraction band on the opposite side. 



4. But if the diameter of a hole be required, then the measure- 

 ment must be made from the outer edge of the outer black diffraction 

 band to the outer edge of outer diffraction band on the opposite side. 

 It must not be forgotten, however, that these rules only apply for a 

 particular focus and a particular adjustment. 



II. The Camera Lucida and its Uses. There are a large number 

 of contrivances devised for the purpose of enabling the observer 

 to see the image of an object projected on a surface upon which he 

 may trace its outlines, but they resolve themselves practically into 

 two kinds, viz. : 



1 . Those intended for use when the microscope is in a horizontal 

 position. 



