694 Transactions of the Society. 



A, very thick and useless, in some places indications of chambers. 



B, scarcely better ; the right valve has possibly a useful thinness, but is not 

 distinct. 



C, has in the upper valve the deceptive image described above with G. 



D, the focus is tolerably good only for the lower valve, where at the same 

 time is visible how the image of the central nodule transverse section passes over 

 into the ordinary transverse section image. 



E, F, G, focus entirely wrong. 



H, shows specially the asymmetry of the secondary rib, exactly as I have 

 described above (fig. 13). Indication of chambers in some places good. 



I, indistinct. 



K and L, two fragments of valve sections, the former showing the chambers 

 very well. 



12. Pleurosigma balticum. The frustule D from the photograph 10. Enlarged 

 from the negative by means of the ordinary portrait objective. Total magnifica- 

 tion, 2,340. The image has been made, fearing that the delicacies of the negative 

 during the printing would suffer considerably, and further to facilitate measure- 

 ments. The valve a' is uninjured throughout, and gives an unblemished picture 

 of the central nodule. The chambers disappear to the eye at some distance 

 from the centre. Instead of the usual membrane-thinning commencing with 

 mid-rib and secondary rib, the wall remains solid throughout (6'), but has two 

 very small projections inwards (c'), between which is a slight depression. The 

 chamber-walls are here distinct, the cavities dark ; the focus of photograph 9 is 

 for this valve more advantageous. The valve a" is uninjured in the lower half, 

 but the two projections c" of the central nodule are still visible ; evidently here 

 has been the extreme fine edge of the section. By comparing photograph 9 with 

 it one detects faint traces of this lower half-valve, which we find again indicated 

 in photograph 12. From the preceding we deduce that this section of the upper 

 half-valve is suflBciently thin for the minutest investigations. The chambers are 

 seen in full clearness and in such accordance with my former diagrams as if this 

 preparation had served for them as model. The girdle-band g b' is not quite 

 distinct, in consequence of being crumbled ; the other g b" partly covered by the 

 section of the frustule E. 



I have formerly given the greatest thickness of the cell-wall as 1'8 )it. Con- 

 sidering what is clear in the valve a' and what is dark in a", and that one must 

 not reckon in a' the dark seam beyond the clear space, we get in this photograph 

 the greatest thickness of the wall near the central nodule as 4* 2 mm., consequently 



4-2 

 the true thickness is ——7: = 1*8 ;u. The height of a chamber lumen may be 



2340 

 estimated at 1/3 of this size ; therefore these are all valves not beyond the power 

 of microscopical observation. 



Valve E a belongs to frustule E, also the girdle-band E gr 6 and the contained 

 portion E i. In the former is seen the indication of chambers tolerably well ; the 

 same may be seen in the valve fragment lying on the left. 



13 and 14. Triceratium Favus. Section No. 11 from a series of 19 vertical 

 sections through a valve. Fig. 13 with high focusing ; fig. 14 with low focusing. 

 Produced with Seibert and Krafft's immersion VII. b ; x 652. The inner surface 

 of the valve is turned downwards. 



15 and 16. Coscinodiscus radiatus. Middle section from a series of 31 vertical 

 sections through a specimen in process of division. Produced like No. 1 ; 

 X 660. Fig. 15, for the greater part of the section correct focus, especially reliable 

 in the centre and a little lower. Here are the T-figures of the chamber-walls 

 distinct, also the four base-membranes. Of less use is the lower end, although 

 pretty distinct. For the upper end the focus is unsuitable. linage 16, gives the 

 correct focus for the upper end, but not so well for the lower end ; the middle is 

 quite unreliable. At both ends is seen the overlapping of the thick girdle- 

 bands. 



17. Triceratium Favus. Thin but not quite plane section. Produced like 1 ; 

 X 680. The section direction is not exactly through the middle of the hexagons, 

 therefore every pair of chamber-walls are brought closer and connected on the 

 outer surface (on the left), and thus between each pair there is a slightly larger 



