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



dark-ground illuminator, but without using a funnel-stop in the 

 lens. Under this method of illumination they are remarkably 

 clear and distinct, and the membrane itself appears to be slightly 

 concave as viewed from the outside of the valve. At one focus 

 and under slightly oblique illumination, one set of bars appears to 

 cross over the other set, as I have represented diagrammatically at 

 fig. 6 ; at this focus the pores are invisible. Upon the specimen 

 from which the fragment is separated both the outer and inner 

 gratings are seen to be composed of hexagonal meshes, as at 

 figs. 7 and 8, and I find it very difficult to get a view of the 

 diamond-shaped meshes on the entire part of this particular 

 specimen, although upon other specimens I have been able to see 

 them and the pores very clearly and easily, as well as the under- 

 lying hexagonal meshes. It would seem as if the outer grating 

 may really be a double structure, with a film of diamond-shaped 

 meshes overlying others that are hexagonal. 



Under certain conditions of illumination a third set of bars can 

 be seen on entire specimens, crossing the diagonals at right angles 

 to the raphe, but I have failed to see any trace of them on the 

 separated fragment represented at fig. 5, so that I think it very 

 probable that they have been dissolved away from that piece, 

 just as also appears to have been the case in the films of 

 P. balticum. For I think there can be no doubt that some such 

 bars exist, because at one focus, under varying conditions of 

 illumination, the gratings appear to be composed of nearly square 

 meshes as represented at fig. 9. At a very slight alteration of 

 focus this appearance alters to the hexagonal one as represented at 

 figs. 7 and 8, which I take to be that of the exact focal plane of 

 the membrane covering the meshes of that particular grating. 

 When the inner grating is examined where the outer grating is 

 stripped off, looking upon it from the outside of the valve, it first 

 presents the appearance of a solid plate of white silex with dark 

 hexagonal perforations in it. At a slightly lower focus this gives 

 place to hexagonal meshes with dark boundaries and the mesh 

 covered with a clear membrane having a pore at its centre ; this 

 latter I look upon as being the true image of the inner grating 

 and the above-mentioned appearance of a white plate with dark 

 perforations as an out-of-focus image produced by some refractive 

 or diffractive property of the membrane, which in some way 

 produces over each mesh a hexagonal shadow. Below the focus 



