290 



CLASSIFICATION OF DIATOMACEjE '. — EUNOTLFJE. 



of occurring in all. — Excluding the family Actlniscece (of whose 

 siliceous skeletons we have an example in Fig. 144, c), which seem 

 to have no adequate title to rank among Diatoms (their true alli- 

 ance being apparently with the Polycyslina), the entire group may 

 be divided into two principal Sections ; one (B) containing those 

 forms in which the valves possess a true central nodule and median 

 longitudinal line (as Pleurosi<jma } Fig. 120, and Gomphonema, 

 Fig. 140, a), and the other (A) including all those in which the 

 valves are destitute of a central nodule (as Surirella, Fig. 129, a). 

 Among the latter, however, we find some (b) in which there is an 

 umbilicus or pseudo-nodule with radiating lines or cellules, whilst 

 there are others (a) which have no central marking whatever. 

 223. Commencing with the last-named division (a), the first 



Fig. 124. 



— """Hfi rir 

 Fig. 124. — Meridion circulare. 



Fig. 125.—fiacillaria paradoxa. 



Family is that of Eunotiece, of which we have already seen a cha- 

 racteristic example in Epithemia targida (Fig. 122). The essential 

 characters of this family consist in the more or less lunate form of 

 the frustules in the lateral view (Fig. 122, b), and in the striae 

 being continuous across the valves without any interruption by a 

 longitudinal line. * In the Genus Eunotia the frustules are free ; 



* The genus Epithemia is specially distinguished by the presence of 

 strongly-marked transverse lines, which were supposed by Prof. W. Smith 

 to indicate canaliculi, but which are regarded by Mr. Kalis (with greater 

 probability) as internal ribs. 



