512 ANIMAL NATURE OF DIATOMIC. 



another point of view, the apparent complication that the 

 simple cell- wall may assume when penetrated by silex. 



Page 22. In the description of Diatomeae it often 

 happens that the valves are confounded with the surfaces. 

 The surfaces are remarkably distinguished by established 

 characters into primary and secondary. In descriptions, 

 it appears to me right to admit, for the sake of brevity, 

 the denomination surface (faccia) for the primary, and 

 side (laid] for the secondary. But the valves do not 

 exactly correspond to the surfaces. The lateral valves 

 are never absent; they sometimes exist solely, as in 

 Pyxidicula arid Podosira. In many cases, principally 

 in Achnanthere and Melosireae, the lateral valves are so 

 bent or curved as to form parts of the primary surfaces. 



But the valves, which exactly correspond with the 

 primary surfaces, are seldom distinct; and perhaps even 

 where there is an acute angle for a boundary, as in 

 Navicula, we cannot say that there is a complete dis- 

 junction. In most cases, the two primary surfaces, and 

 therefore the corresponding valves, unite into a continuous 

 plane, with a re-entering curve. The distinction of ter- 

 minal surfaces, very evident, especially in the SurirelleaB, 

 is most important. When there are distinct terminal 

 surfaces, the primary surfaces are very limited. But 

 these terminal surfaces may belong to the primary valves, 

 or the secondary, or lateral. If we take the Eunotiea?, it 

 seems evident that the more or less acute or obtuse 

 extremities belong to the lateral valves, and therefore 

 the two primary remain entirely separated. But in the 

 Surirelleas, these seem again to belong to the primaries, 

 and to participate with them in the process of de- 

 duplication. 



Page 23. The solid products which, under the form 

 of incomplete diaphragms, project into the inner cavities 

 of Eiddulplda and Terpsinoe, belong to the lateral valves. 

 I can say nothing of Ctimacosp/Henia, but in the Isthmia 

 nervosawe have something similar, though less developed. 

 I think we may also refer the septa of Actinoptychus to 



