ANIMAL NATURE OF DIATOM&&. 485 



correspondence with the cells above mentioned. The 

 lateral processes are more or less completely denuded of 

 the fine cellular membrane, and are united in con- 

 catenation with each other by a cushion (cuscinetto), 

 bounded externally by a precise line, but internally 

 irregular. 



The deduplication takes place in this genus as in those 

 preceding. The double ligament merits observation, 

 which remains to unite the median lobe of both surfaces 

 of each extremity with the corresponding (ligament) of the 

 adjoining frustule. This ligament, from the gradual elon- 

 gation, breaks across in the middle, maintaining its rigidity. 



In the Adriatic we have the two species, qui?iquelocu- 

 laris and trilocularis. To the second I refer a form with 

 smaller and less elongated frustules, which at first I 

 thought a new species, and as such communicated it to 

 Kutzing, who, on the other hand, believed that it belonged 

 to the quinquelocularis. I note this circumstance to in- 

 dicate an important organic condition. When the frus- 

 tules are empty and diaphanous, we can easily see, 

 owing to the transparency, the costae of the surface 

 opposite to the one observed, and especially the two 

 extreme ones, which correspond to the least thickness, 

 and can therefore be seen along with those of the anterior 

 surface. The number of the so-called cells is then appa- 

 rently increased by two. It happens in this way that 

 the quinquelocularis sometimes resembles the septem- 

 locularis. 



Kutzing ascribes to this genus, as a doubtful species, 

 the Denticella Fragillaria, of Ehrenberg. 



In the quinquelocularis, I once saw a frustule broken 

 in the direction of the conjunction of the lateral valves 

 with a median cincture, and surrounded by 4 6 poly- 

 hedral corpuscules, of a diameter equal to one fourth 

 that of the frustule, and of a cellular structure, adhering 

 to the margin of the fracture. 



Finally, we cannot but remember the Isthmia ner- 

 vosa, which seems to differ from Biddulphia in no 



