38 GENEEAL HISTORY OF TfiH INFUSOEIA. 



the silicioiis sMeld or lorica is four-sided, and composed of four pieces or 

 valves. Although this appears to be the true stnictm-e of some species, omng 

 to the ready separation of the connecting membrane into two portions, yet the 

 majority oifer no countenance to the notion, the connecting membrane 

 forming a continuous oval or circular ring. In Triceratium, however, is an 

 example of an even more pseudo-multiple composition ; for its prismatic tri- 

 angular frustule breaks up into " two triangular plates or walls of silex 

 forming the ends, and into three oblong rectangular pieces or bands forming 

 the three sides, the latter usually dividing themselves into several elongated 

 paralleliform pieces" (Brlghtivell, J. M. S. i. 248). 



Again, in several genera, doubtfully arranged by Ehrenberg among the 

 siHcious Bacillaria, e. g. Dictyocha (XII. 62, 63) and Mesocena (XII. 71), the 

 individuals are represented as composed of several segments united together. 

 Each valve consists of a sihcious lamina supei-posed on an organic soft 

 lining (or primordial) membrane which immediately encloses the contents of 

 the cell. Nageli speaks of a mucilaginous peUicle on the inside of the 

 organic layer as a sort of third tunic ; and Kiitzing likewise discovers a thin 

 stratum brought into view when recent frustules are dried, and particularly 

 after heating them to redness, in the shape of an opake bro^vnish stain, or 

 of brown hues or points, extending not unfrequently over a considerable 

 portion. To this supposed independent material its observ^er applied the 

 name ' cement/ imagining it to be the connecting matter of the valves and 

 of frustules when in luiion, and attributed its bro^Ti colom^ to the presence 

 of ii'on. This presumed layer of cement we can regard as nothing more than 

 the stain produced by the oxidation of the salts of iron in chemical union 

 ■with the silica, as Prof. Erankland has shown (p. 37). However, Meneghini 

 adopts the notion of this third envelope or cement, inasmuch as he observes 

 it to be constant, without employing the means used by Kiitzing to display 

 it, not merety in the species enumerated by its discoverer, but in many 

 others, and possibly in all (R. S. 1853, p. 361) : — " Por to me (continues the 

 same author) it appears to correspond with that fine membrane of the Ach- 

 nanthidia, which, according to Kiitzing's o^^ti observation, is always \isible 

 whenever the two new individuals (into which every Diatom is resolved in 

 its multiplication by deduplication) begin to separate. The lines and points 

 supposed to belong to the subjacent shield belong very frequently to this kind 

 of covering." The analogy expressed in the quotation just given, between 

 the delicate stratum — the ' cement ' of Kiitzing, and the secretion poured 

 out when self-division is proceeding, we cannot regard as correct ; for this 

 latter is a special and usually not persistent coating, in all j)robability exuded 

 through the fissui^es or pores uncovered by the silieious lamina, by the sub- 

 jacent organic membrane, and is T\dthal destroyed by the heat generally 

 required to brmg the ' cement ' into view, whereas the presumed coat is 

 represented to be constant and also permanent both under the operation of 

 fire and of acids. However, the belief in the existence of a vegetable mem- 

 brane outside the silieious epiderm gains ground ; for Mr. Shadbolt, in his 

 presidential address before the Microscopical Society, 1858 {T. M. S. 1858, 

 p. 72), states it as the result of his researches, that the frustule of Arach- 

 noidiscus and of other forms consists of a silieious frametvorJc, over which is 

 stretched a species of membrane, whether silieious or not he does not presume 

 to decide, but certainly pliant to a considerable extent, capable of being par- 

 tially rolled up by mechanical agency without breaking, and elastic enough 

 to return to its original position when the extraneous force is removed. 

 " The structiu-e noticed by Mr. Eoper in Coscinodiscus Jahyrinthus, and by 

 myself in the more common species C. radiatus and Triceratium favxis, I 



