486 Miscellamoies. 



place of a mirror, be provided witli a reflecting prism, the object may 

 be placed opposite the side external to the microscope. The instru- 

 ments best adapted for the manipulation wliich we are describing are, 

 however, those whose illuminating apparatus consists of a mirror and 

 converging lens, which can be shifted up or down. The lens being 

 removed from the ring which supports it, the object is substituted in 

 its place. The relative magnitudes of object and air-bell must be 

 such that the image shall be exceedingly minute when the object is 

 tolerably near to the stage. On afterwards increasing the distance 

 between the object and air-bell, it is not difficult to find the limit at 

 which the image (under a given magnifying power) is barely visible. 



Of course it is impossible to measure directly the dimensions of 

 this most minute visible image, for our best micrometric methods 

 will here be found of no avail. Yet their size may be estimated with 

 extreme accuracy in the following manner. At the same distance 

 from the air-bell and in place of the object used, substitute another 

 body, such as a piece of card, of 4 to .5 centimetres = If ths to 

 inches diameter, which has been exactly measured. Let this be now 

 again measured just as if it were a real object. By dividing the real 

 diameter by the apparent diameter, the amount of diminution is found ; 

 and this is the same for all objects at alike distance from the air-bell. 

 We have, consequently, nothing to do, in order to find the amount 

 of diminution of the image of the more minute object, but to divide 

 its true diameter by the figure expressing the diminishing power. 



For example, let the true diameter of the greater object be 5 centi- 

 metres =to 1'969 English inch, and the diameter of its image = 

 ?2-2 micromillimetres*, = -00l27 Enghsh inch, then the figure ex- 

 pressing the amount of diminution will be .^^j^\=1553 very nearly. 

 If now the smaller object have a diameter of 1 75 micromillimetres 

 = •00689 English inch, then must its image at the limit of vision 

 be in diameter =' ^^%|%9 = -0000044, or about ^-^i^^th of an English 

 inch. When exact micrometric methods are employed, it is easy 

 in this way to estimate the diameter of an image even to millionth 

 parts of a millimetre, i. e. to 25,400,000th parts of an inch. 



As for the object suitable for these investigations, it is plain that 

 we have an extensive choice. To find the limit of vision for bodies 

 of a round or long thread-like form, grains of pearl sago, or vegetable 

 bodies, such as mustard-seed or the pollen-granules of many plants, 

 hairs of animals, metallic wires, &c., may be employed. Small round 

 openings and chinks may serve for the determination of the visibility 

 of positive images of light. In the last case care must of course be 

 taken, by means of suitable screens, to shut off all light except what 

 passes through the aperture. To determine the defining power, 

 metallic wire-gauze is a suitable object, or two holes placed near each 

 other in a black metallic plate. The images of such objects resemble 

 exactly a double star viewed through a telescope (Jiijker). The 

 bodies may likewise be placed in different circumstances in order to 



* The micromillimetre is equal j^Vw millimetres= '0000394 English 

 inch. 



