( 851 ) 



XX. — Researches on the Structure of the Cell-walls of Diatoms — 

 Eupodiscus* By Dr. J. H. L. Flogel. 



( Read 12th Decemher, 1883.) 



Amongst the specimens of Triceratium received from Herr Moller, 

 Eupodiscus argus Ehrenb. is well represented, the structure and 

 sculpturing of which I have endeavoured to determine by sections. 

 Eesearches were formerly made by Slack (25), who considered the 

 membrane to consist of minute spherules. The different appear- 

 ances this form presents in reflected light as compared with other 

 diatoms, and which point to quite different sculpture details, have 

 been discussed by J. Deby (2, p. 13). Indeed, the sections coufirm 

 this, presenting such a peculiar image that it is difficult to institute 

 a comparison with others. The spine of an Eupodiscus valve has 

 been described and figured by 0. Mtiller (15, p. 633, pi. xv., 

 fig. 8). Wells asserts that he has seen irregular depressions on the 

 valve, which are closed by a membrane at the base and divided by 

 lines which are coarse at the edge (29). 



I send the photograph of a coarse section, being No. 9 of a 

 series of nineteen sections through a valve, which at the same time 

 contains one of the three spines (photograph 8). Only very 

 indistinctly we see in some places that the mass must consist of 

 nearly symmetrical chambers resembling closely in their general 

 outlices the Triceratium chambers. If we compare with this the 

 thinnest sections — I have given a drawing of a jDortion from such a 

 section in fig. 144 — it becomes very difficult to trace the chamber- 



FiG. 144. 



\Mjti>^:^^sj 



like spaces. Sometimes I see a delicate line like the vertical wall 

 of a Triceratium, but in most instances this line is so darkened by 

 what seems a small granular mass of high refracting power that 

 one loses trace of it. I consider this granular mass to consist of 

 numerous excrescences from the chamber-walls, without however 

 being able to give proof for my assertion. Here then is a new field 



♦ In consequence of omiesions in the (ext this could not be printed in ita 

 proper place ante, p. G72. 



