The Structure of Diatoms. By Dr. J. H. L. Flogel. 673 



projection of lines on the inner side of the membrane, whilst the 

 outer side is perfectly smooth. It can be seen more distinctly in 

 sections, but in proceeding we shall soon have to admit that a part 

 of what we stated has to be reconsidered. The following relates 

 only to Achnanthes hrevipes Ag.* 



§ 1. "We have to distinguish three sorts of markings in 

 Achnanthes. This diatom has an unsymmetrical shape at the 

 division-plane, so that only the two middle sections vertical to this 

 plane give symmetrical halves. Therefore we get: — (1) A dorsal 

 valve characterized by greater convexity ; it has a mid-rib without 

 nodule running evenly from one end to the other ; close to it 

 run the smooth transverse lines, the interspaces of which are wider 

 than those of the ventral valve ; between every pair of transverse 

 lines are seen in most instances three rows of dots, sometimes only 

 two. (2) A ventral valve, characterized by a thick depression in 

 the middle ; it has a mid-rib with even strias on both sides, in the 

 centre a large nodule which at right angles to the mid-rib extends to 

 the edge, thereby producing, the same as with Stauroneis, the 

 image of a cross. The transverse lines are finer than in the dorsal 

 valve; between each pair, as a rule, one row of dots, sometimes 

 two. (3) The girdle-band, always with delicate strife vertical to 

 the division-plane, which however is subject to variations as we 

 shall presently see. The figs. 33, 37, 38, and 39, plate X., of 

 A. hrevipes sufficiently illustrate this description. 



Two good serial sections are obtained, running vertical to the 

 division-plane and to the two mid-ribs (in fig. 33 this is delineated 

 by lines 1-3) ; the third series 1 made approximately parallel to 

 the division-plane. My attempts in the third direction of space 

 (the horizontal) failed. From one of the two former series, 

 numbering twenty-three sections, I have delineated three, viz. 1, 5, 

 and 15, figs. 40-2, plate XL No. 1 is a marginal cut; 23 has 

 nearly the same appearance. No. 15 cannot be far from the middle, 

 because among the succeeding numbers are a few of too great 

 thick' ness. 



Examining first the general form, we see from the sections 

 in which all three conform, that the ventral valve is depressed, 

 trough-like, along the mid-rib, whilst the dorsal valve appears half- 

 cylindrical, that is to say, rounded ofi" convexly. By this feature 



* The difference between A. brevipes and longipes Ag. is often very great. 

 A. I'jn/jipes commences witli a short pedicle ; the pointing or rounding of the valves 

 is somewhat variul>lo, and tti(i distance between the stri.'o is not always definite. 

 From tlie stria distances and the length of the pedicle, I believe I liave deter- 

 mined the Bpecimens investigated to bo A. brevipes, but tliey might belong to 

 A. lorujipes. The specimen figured by Pfitzer as A. brevipes (I'J, pi. vi. lig. 15 s) 

 I should rather suppose to be A. ventricosa Ktz. ; anyhow, this variety is not 

 the one investigated by me under the name of A. brevipes. I have found 

 A. ventricosa on the sea-shore near Sylt, but could not make use of it in tho 

 present investigation {vide 5, p. 737). 



