254 PHOTOMICROGRAPHY 



lyra, Surirella gemma, etc., is fairly constant, the beginner 

 has the advantage of being able to compare his results 

 with figures and descriptions published by other workers, 

 and as test objects they are of the greatest value. 



Critical and monochromatic illumination, and objec- 

 tives of great numerical aperture and perfect corrections 

 are needed to give the best results. Apochromatic 

 objectives are very desirable, but good results are to be 

 obtained with achromats and suitable screens. The most 

 useful powers are J", J", and iV". The colour filters 

 should transmit a light approximately monochromatic, 

 and of small wave length to increase resolution as much 

 as possible. On account of the difficulty of focussing with 

 blue light, green screens are easier to use, and achromatic 

 objectives generally work to better advantage with green 

 light. In certain cases, owing to insufficient contrast, it 

 is impossible to get all detail brought out by axial light, 

 and oblique illumination must be resorted to, but great 

 care must be exercised in its use, and it should only be 

 employed to show up in greater contrast structures that 

 can already be made out by axial light. The excessive 

 use of oblique light tends to the formation of false 

 images, and it is no uncommon thing to be able to make 

 out and photograph dots or striae lying outside the diatom, 

 which must, of course, be purely optical phenomena. 

 When using objectives of high numerical aperture it is 

 often necessary to cut down the aperture slightly by 

 means of the iris diaphragm of the condenser, in order to 

 prevent flooding of light and to get contrast. Great 

 care must be taken in closing down the iris as diffraction 

 effects are so easily produced (Plate 12, Fig. D). 



It is advisable to commence work on one of the large, 

 flat diatoms, such, for example, as Arachnoidiscus 

 (Plate 39. A), and to take the photograph with a low 

 objective and long camera extension ; there will be no great 

 resolution, but such a diatom calls more for flatness of 



