The Structure of Diatoms. By Dr. J. H. L. Flbgel 693 



graphed a few weeks after. It is No. 7G out of 150 transverse sections. Through 

 a number of 14-lG frustults, partly parallel, partly one on the other. Image 9 

 Wiis taken with somewhat higher focus, and image 10 with lower focus; the 

 ditierence is naturally very insignificant. For the examination of these two 

 images and of phot'^giaph 11a lens shoiUd be used. For the representation of 

 the sculpture my former description of the transverse sections of P. halticum 

 should be referred to (6, pp. 481-5, figs. 13, 14 and 15). The separate frustules 

 are marked A-L. For the compreliensiou of the arrangement of the frustules in 

 comparison with photograpli 11, we must observe that only the Fleurosigmt A to 

 D are in the original position, that the gum section on the right of D on account 

 of extreme thinness is broken ofl', and that this broken-off and still further broken- 

 up poition has fallen obliquely across the other portion which remained intact. 

 The thinness of the gum-chip in this instance can be best estimated by noting 

 that hardly any trace of the collodion limit lines can be seen {vide photograph II), 

 which miist have existed below B and D. About the separate sections the 

 following is to be observed : — 



A, very thick section, kidney-shaped within. The left valve is broken in the 

 middle, only the one half is sten. The right valve is broken up, but both the 

 chief and secondary ribs, with a portion (jf thin membrane attached to the former 

 distinctly visible; the chambers slightly indicated. 



B, a slightly thinner section, strongly notched witl.in. The left valve twice 

 fractured; the right valve quite uninjured; only well focused in image 10; the 

 t.iin membrane-stria next to the mil-rib very beautiful; slight indication of 

 chambers ; the girdle-band at the top is in normal position, below out of place, in 

 ph-.togruph 10 clearly duplex. 



C, section thinner than B. Content wanting. The upper valve uninjured. 

 Best focus in photograph 9 ; indication of chambers very good, particularly iu the 

 left half; a depressed furrow beside the raid-rib good, also the secondary rib. 

 The luwer valve is broken iu the thin portion, although both halves are not much 

 displaced. Focus in both images not exact. The left girdle-band is injured ; on 

 the right clearly duplex; both still adhere to their original valve, altiiough 

 slightly disjilaced one towards the other. 



D, "this magnificent central noJule section, which forms the chief object of 

 representation, and is rendered on a larger Sv.ale in photograpli 12, I shall describe 

 further on. 



E, injured section, only the lower valve good. The content adheres to the 

 girdle-band. On the left a few fragments of valve transverse sections. 



F, a much injured section. The upper valve broken in two pieces, but the 

 fracture is not in the thin membrane stria; the half lying a little higher shows 

 in photograph 10 veiy good indications of chambers ; the lower half to which is 

 atta'hed the content is not so good ; tolerable in photograph 9. The lower valve 

 broken up ; the ginlle-band has got out of place. 



G, imperfect section, which very ckarly proves whiit kinds of deception can 

 take place in researches not conducted critically. In the half-valve on the right 

 at the top one sees in photograph 9, without much trouble, the outlines of a base 

 membrane, with ridge elevations noded at the ends. A little more imagination 

 will add outer openings of the chambers. Photograph 10 destroys all these 

 illusions, and proves at the same time that this section is much too thick tor such 

 studies ; it is almost exactly the reduced copy of Finnularia, photograph 2 at r 

 {vide supi-a'). 



11. I'icurosiijmx balticum. Everything like 9 and 10, except the transverse 

 section No. 80 out of the same seiies (Nos. 77-8-9 were useless fragments 

 cut through totally different portions of the bundle). This image chiefly 

 serves U) shcjw that by using my section method the separate frustules can be 

 identified from section to section, whereby real series of sections can be obtained. 

 A to D can be easily recfjgnized by comparing with images 9 and 10. Below 

 B D E run two slightly bent parallel lines through the field of view ; these form 

 the limit of the trauoversely cut delicate collodion film on which the arrangement 

 of the Plenr'MifjiiM took place. On the right of E the section is broken off and 

 another gun.-chip has fallen obliquely on it ; the frustule sections H and I 

 therefore lie one ujjon the other. For tiic separate sections the following remarks 

 may l>e serviceable : — 



