676 Transactions of the Society. 



by its peculiar regular fine pearl-like sculpture, figs. 45 and 52. 

 Now, what are these pearl rows ? I believe they must be inter- 

 preted similarly to the sections of the Flensburg Fleurosigma 

 described by me (6, pp. 475-8), hence, in this case, probably a long, 

 extended, cyHndrical chamber within the membrane. The apparent 

 projection on both sides of the surface must be an optical effect. 

 Clearness in these details with the extreme delicacy of the object 

 is hardly to be expected. The central nodule of the ventral valve 

 is a strong inward-projecting thickening, without any other dis- 

 tinction except that from every side it sends off a thinner line in 

 the transverse direction to the edge of the valve. 



§ 2. With this we conclude our examination of sections in 

 general. I believe I have thrown some new light on the com- 

 plicated structural details of the species Achnanthes, although I 

 admit that much remains to be done. With the ample material at 

 hand it became interesting, apart from the above results, to make 

 the attempt to discover the development-processes which take place 

 in the gradual formation of these sculptures during ordinary 

 fission. Now that we know more exactly the connection of the 

 girdle-band with the valves, the former, not being structureless like 

 most of the fresh-water diatoms, will probably furnish data for 

 further ehicidation of these hitherto obscure problems. In aid of 

 further research we should also avail ourselves of simple surface- 

 views as well as the examination of numerous freshly imbedded 

 and well-preserved specimens in balsam, because with the help of 

 transverse images the appearances can be correctly explained. I 

 do not hesitate to add here the observations which I have made 

 with a larger number of balsam specimens of Achnanthes hrevipes 

 even at the risk of engaging in controversy. 



Taking a recently divided specimen, such as is shown in 

 fig. 33, and examining it with reference to the formation of the 

 girdle-band, one finds that between the two valves, viz. between 

 the one older and the one younger, there is no space. We observe 

 the marginal line, and close to this extends the young valve. If we 

 keep well in mind the image of the marginal line of another fuller 

 grown specimen, we shall be easily convinced that the younger 

 specimen has only one single line. Fig. 33 shows the frustule at 

 the edge where the young valves have only recently obtained the 

 necessary solidity to enable them to withstand the influence of 

 contracting fluids (I might have started from earlier stages, but my 

 objects not hardened with osmium show all sorts of bendings of the 

 young valves which I attribute to the modus operandi). Adjust- 

 ing the left cell as the optical middle section one sees the edge-line 

 of the old dorsal valve like projecting nodules : close adjacent is a 

 smaller nodule which can be nothing else than the commencement 

 of formation of the marginal line of the younger ventral valve. On 



