Royal Microscopical Socitty. 239 



impairing the sensitiveness of the eye. There is no possibiUty of 

 confining the light exdusivehj to the object to be viewed, which is an 

 important point. In this last respect my immersion paraboloid is 

 not better, and also requires the bull's-eye for parallel rays. 



The diagram, five times the size of the original (see p. 240), 

 illustrates the plan that I have arranged to overcome these defects, 

 a is a cylinder of glass half an inch long and four-tenths in diameter, 

 the lower convex surface of which is polished to a radius of four-tenths. 

 The top is flat and polished. Starting from the bottom edge, the 

 cylinder is worked off to a poKshed face at an angle of 64° : close 

 beneath the cylinder is set a plano-convex lens of 1 j focus. Parallel 

 rays sent through the lens, after leaving the lower convex surface 

 of the cylinder, would be refracted to the point shown by the 

 dotted lines if continued in solid glass, but by impinging on the 

 inclined polished surface (which is far within the angle of total 

 reflexion) they are thrown on the flat segmental top ; here they 

 would be totally reflected and beaten down again to the point, b, 

 outside the cylinder ; but if an object-slide, c, be laid over the flat top 

 with an intervening film of water, the rays proceed on till the focal 

 point reaches the upper surface or is slightly beyond it ; here total 

 reflexion now takes place ; all the light is concentrated to a minute 

 spot in the centre of the field of view of the microscope, and most 

 of the rays are available for any object brought there by tra- 

 versing the slides over the water top of the illuminator, which must 

 be kept full without allowing any to run down the reflecting sur- 

 face. It will be seen, in order to get the focal point in the centre 

 of the microscope, that the lens centre must be excentric, but this 

 does not involve the slightest inconvenience, as the excentricity only 

 amounts to a little over two- tenths of an inch, and is so small that 

 the same adjustment of the mirror serves during an entire revolu- 

 tion. The apparatus rotates on the focus as a centre. The 

 management of this illuminator is very easy and simple ; its fitting 

 goes into the ordinary sub-stage, and has an independent rotary 

 movement of its own, like that of Nachet's prism. The cylinder is 

 brought uj) nearly level with the stage. The centre of rotation is 

 set true by a dot on the fitting seen with a 1^ object-glass. A drop 

 of water is then placed on the top, upon which the slide is laid. 

 The required objects on the slide are found by a low power, and 

 may be distinguished by their brilliant appearance, while those on 

 the cover are nearly invisible. The light is thrown up by either 

 the plane or concave mirror. The former is generally the best and 

 most controllable. The lamp should not be placed much beyond 

 the stage, else its direct rays will get underneath and mar the 

 blackness of the field. Having got the best effect, say on a diatom^ 

 or insect scale, by tilting the mirror, we now proceed to rotate the 

 illuminator. During this the most exquisite unfolding of structure 



e 9 



