90 



Bacillariew 



which have been most carefully examined with this object in view have 

 failed to reveal any trace of true perforations through the cell-wall. They 

 occur beyond question, however, in many species, and Schiitt states that it is 

 highly probable that they exist in a large number of others. Schiitt instituted 

 a distinction between pores and dots; and O. Mliller ('00) has still further 

 emphasized this distinction by suggesting that all the minute circular dots 

 over 0'6 fju in diameter, which are tiny cavities resembling pores but not 

 actual perforations, should be termed ' poroids ' in contrast to the true pores 

 which vary from O'l fi to 0'5/-t in diameter. 



Both O. Miiller and Lauterborn have shown that in some of the larger 

 naviculoid diatoms the cell-wall is destitute of 

 pores and is only broken through by the cleft of the 

 raphe (see p. 92). In the large species of Navicula, 

 of the section Pinnularia, such as N. nobilis, N. 

 major, etc., the bilaterally arranged costse on each 

 valve are elongated chambers on the inner side of 

 the cell- wall. Each of these furrow-like chambers 

 communicates with the interior of the cell by a 

 more or less wide opening, and the edges of the 

 openings appear in the valve-view as fine longitu- 

 dinal lines crossing the costse (fig. 57 B). It 

 must not be assumed, however, that pores are 

 absent from the walls of all those diatoms which 

 possess a raphe, because some species of Gyrosigma 

 (= Pleurosigma), and also JEpithemia Hyndmanni, 

 possess both a raphe and fine pores, and it is 

 reasonable to suppose that a similar structure of 

 the wall occurs in other diatoms. 



In Eupodiscus Argus, which is a frequent 

 marine diatom, there are numerous cup-shaped 

 chambers opening widely to the exterior, the walls 



o f these chambers forming an areolar network of 



. ., 

 ridges on the surface of the valve. The walls of 



the Cambers are finely papillate, and several 



minute, obliquely sloping canals perforate the base 

 of each chamber (fig. 62 D). 



The structure of Triceratium Favus is some- 



what similar, only the chambers opening to the exterior are polygonal and 

 their walls are smooth. Minute poroids occur on the inner side of the inner 

 walls of these chambers, and pore-canals pass through the flange of the valve 

 as minute tubes (fig. 61 B and D). 



In Isthmia nervosa there are larger primary and smaller secondary 



B 



Fig. 63. Stephanopyxis Pal- 

 meriana (Grev.) Grun. to 

 show the spines on the valves 

 by which the cells are united 



pairs of these spines to 

 show their tubular character. 

 (After 0. Miiller.) 



