N. E. BROWN ON THE STRUCTURE OF DIATOMS. 321 



Vol. III. p. 306, regards the valve as "composed of two layers 

 of grating"; whilst Mr. E. M. Nelson in the Journ. Q.M.C., 

 Ser. 2, Vol. XII. p. 99, fig. 4, states that " in P. balticum 

 and allied forms the upper membrane has slit-like apertures in 

 longitudinal rows, while the lower membrane has circular 

 apertures (fig. 4), where the circular apertures in the lower 

 membrane are seen through the intercostal silex of the upper 

 membrane and in a line between the slits." Finally we have 

 Mr. O'Donohoe's interpretation referred to above. 



Until August 1913 I held the view (which I think is the 

 prevailing one) that the black dots visible on the valve of a 

 diatom were pores or perforations passing completely through 

 its substance, and that the white-dot view was an out-of-focus 

 one. Now, however, the examination of Mr. Capell's slide has 

 demonstrated to me and to others who have examined it with 

 me, conclusively and beyond any room for doubt, that many 

 (possibly all) of the black dots that are ordinarily seen on a- 

 diatom are not pores at all, or at the most are only pits con- 

 taining the pore-bearing membrane, and that the white-dot view 

 is often much more correct for seeing what I believe to be the 

 true pore-structure than has been supposed. 



I have long been puzzled at the behaviour of black dots under 

 high magnification, and have therefore suspected that they 

 were not quite what they seemed to be for some time past, but 

 I think the evidence of Mr. Capell's slide fully explains their 

 nature. 



In any perfect valve of P. balticum it is easy to obtain a view 

 of a grating-like structure with square meshes, formed of bars- 

 or rods of silex crossing one another at right angles. In the 

 partly dissolved films on Mr. Capell's slide this grating is not 

 evident, but instead the films are seen to consist of parallel 

 dark rods having a beaded appearance, held in place by a 

 membrane of silex (see Mr. O'Donohoe's figures, op. cit., t. 14 r 

 figs. 3 and 4). At the ends or other parts the rods are seen 

 to project in a ragged manner. These rods are those which lie 

 parallel to the raphe in the perfect grating, while those which 

 in the perfect diatom form the transverse bars of the grating 

 structure have been dissolved, leaving no trace, or only a very 

 faint one, visible. When examined with an oil-immersion objective 

 at a magnification of 2,000 to 3,000 diameters these bars are 



