N. E. BROWN ON THE STRUCTURE OF DIATOMS. 329 



tests ; with lenses having a smaller aperture than about 68 N. A. 

 only the primary bars of the framework are visible ; with lenses 

 of a larger aperture, the secondary bars (i.e. those of the cell- wall) 

 become manifest as very fine lines between the primary bars; 

 finally with lenses of large aperture and at a magnification of not 

 less than 1,800 diameters these fine lines or bars are seen to b9 

 connected by finer transverse bars so as to form a ladder-like 

 structure, with a minute pore at the centre of each bead-like 

 space formed by the cross bars, or at another focus the bars can 

 be resolved into the appearance of rows of dots. 



Navicula serians. The structure of the valve of this species 

 seems rather difficult to understand. When I first examined it 

 in search of pores, I found it had a rather coarse grating, with 

 oblong meshes arranged in six to seven rows on each side of the 

 raphe, the longer diameter of the meshes being transverse to the 

 latter. These meshes are closed by a very thin membrane of 

 silex, at the centre of which can be seen, at a magnification of 

 3,000 diameters, a minute dark dot, as represented at the upper 

 part of fig. 10. This clot I take to be a pore. With central 

 light only a very faint indication of it is seen ; but when a small 

 central stop is placed in the condenser it becomes clearly visible. 

 This structure is all that I at first noted. But having re- 

 examined this diatom with great care under all conditions of 

 illumination at my command, I have detected structure which had 

 previously entirely escaped my notica. For I find that if the 

 outer surface of the valve is illuminated by a Leitz dark-ground 

 illuminator and examined at a magnification of not less than 

 2,000 diameters, without reducing the N.A. by using a funnel- 

 stop, a second grating exterior to and superposed ^upon that 

 above described can be distinctly seen. This outer grating is 

 evidently extremely transparent and practically invisible by 

 central light, so that it very easily escapes notice. I have found 

 that the easiest way to make it evident is, first to get [the mem- 

 brane of the coarse meshes in focus, as represented at the upper 

 part of fig. 10, then gradually but very slightly raise the lens 

 above that focal plane, until two dark dots appear over each 

 mesh. If these dots are very accurately focused and the dark- 

 ground illuminator manipulated so as to illuminate the diatom 

 with light reflected upon it from the under surface of the cover- 

 glass, the surface of the valve will be found to have the 'appear- 



