ANIMAL NATURE OP DIAT0MEJ3. 457 



which are considered to be canals. I have not been 

 able to see the median aperture, which Kiitzing describes 

 and figures in some species {lielmintosum). As to the 

 distribution of the internal substance, though nothing 

 positive can be deduced from the observation of specimens 

 dried and softened, or even preserved in. alcohol, it is 

 yet remarkable that even in such a state it presents 

 different conditions that are constant in each species. In 

 most instances it is collected along the central part of 

 the lateral surfaces, so that it exposes a colourless longi- 

 tudinal area in the centre of the secondary surfaces, and 

 only the two extremities of the primary surfaces are 

 colourless. Sometimes, too, the median line of these last 

 is colourless, because the colouring matter is lodged 

 {niccJiiata) in the four inner angles. It is not unusual 

 to find it condensed in a single central globe. In a 

 few species only I have seen it constantly divided into 

 two portions corresponding to the extremities, whilst the 

 median area of the entire body of the Naviculse remained 

 un coloured. 



44. DiCKiEiA. — Tliycoma foliaceum {phyllomd) basi 

 substipitatum. Naviculce in membrana gelinea irregu- 

 lariter sparsce. 



Kiitzing does not describe the structure of the so-called 

 membrane of the only species {D. ulvacea) of this genus, 

 and I am sorry that I cannot consult Berkeley's descrip- 

 tion. This membrane may be formed of cells including 

 the Naviculse, as in FrusMice, or of elongated cells or 

 tubes, as in Schizonemce, or of a single capacious flattened 

 cell comparable to that in Monnemce ; or, finally, this so- 

 called gelatinous substance may be a mass entirely con- 

 tinuous, in which the Naviculse may be immersed, as 

 indicated by Kiitzing " in der gallertartigen Haut einge- 

 bettete," as we have suspected of some species regarding 

 which there seems to be a doubt whether they belong to 

 the MonnenuB or the Schizonema. 



The last seven genera, {FrustuKa, Berkeley a. Rapid- 



