The structure of Diatoms. Bij Br. J. H. L. Fldgel 511 



diatoms there was a glutinous organic substance which could only 

 be protoplasm. In 1870, Prof. Dippel's work (3) appeared refuting 

 Schultze's view, and drawing attention to the fact that the mid- 

 rib is not a cleft but rather a thickened line, having on the sides 

 narrow longitudinal strias without perforations.* Then came 

 Pfitzer's work (19), who discovered the long-searched-for cleft and 

 rerpresented it so well that its existence is no longer doubted. He 

 declares the detection of this cleft to be a task to be solved only 

 with the highest powers, and draws a figure of a very narrow 

 V-shaped opening through the thick envelope. Under existing 

 circumstances it was of great interest to test Pfitzer's statements. 

 The conclusion I came to was to agree entirely, without reserve, 

 with Dippel, and I must therefore deny the perforation of the 

 cell- wall. This result cannot, however, be easily obtained. The 

 number of more or less good transverse sections of Finnularia 

 I have is about 600 ; in most of them I actually observed in the 

 midrib a fine transverse cleft. Its direction and its fineness are 

 shown in figs. 1-3. They can be well seen with 300-400 where 

 the section is good. With regard to direction I rarely see the 

 cleft so Y-shaped as drawn by Pfitzer ; on the contrary, in most 

 instances it commences on the outer surface at right angles to it. 

 Then arise a great variety of changes in the appearance. We 

 meet with sections in which it goes straight to the inner surface ; 

 others where the vertical portion has a small hook turned 

 inwards ; again, others where the portion facing the inner surface 

 is obtuse to the vertical, — this case (not at all uncommon) is repre- 

 sented in fig. 3. Such a change may be explained by a real 

 difierence in the object, which Schumann also quite correctly found 

 with the surface view ('23, p. 73). A careful comparison of a great 

 number of transverse sections, made according to method 2, will 

 show in most cases the cleft going up to the inner side of the 

 membrane. The question remains very doubtful whether the base 

 is closed by a very thin envelope. After having made collodion 

 casts, the image of the outer surface could be easily inter- 

 preted ; the fine cleft became filled with collodion, and in harden- 

 ing a distinct midrib remains, almost exactly the same image as 

 we see in the surface view. But the inner surface remained quite 

 obscure. If a perforation of the membrane really exists, a similar 

 midrib should be seen ; but this is not the case. I have made 

 a considerable number of such impressions, and not the slightest 

 trace of a midrib could be seen, even with oblique light. 



These impressions fully convinced me that the fine cleft was 

 closed at the base. That the transverse sections mostly show a 

 fully developed perforation may be attributed to the following 



* TLia is what I proved simultaneoubly with Pkuroskjma (stiiie without 

 Bporce). 



