OF THE DIATOMEiE, 47 



membrane in Pleurosigmn is a remarkable feature of that genus, whilst in 

 Gomplionema (XII. 28, 53), and other genera with cuneate (wedge-shaped) 

 frustules, the figure is due to the greater development of this segment at one 

 end than at the other. In Amphiprora (XIII. 5, 6), Achnanilies, Himanti- 

 clium (XII. 50, 52), and Melosira, the connecting membrane is striated, and 

 in Biddiilpliia (II. 48), Isthmia (X. 183), and Amphitetras (XI. 21, 22) is 

 cellulate or areolate. 



In certain genera the connecting membrane takes on an extraordinary 

 development, which greatly modifies the figure of the frustules. Instead of 

 being Hmited to the interspace between the opposed valves, it extends on 

 either side beyond the sutures (XII. 9), presents itself as a band of greater 

 or less width, and acquires an unusual persistence. Under this form it con- 

 stitutes the ' cingulum ' of descriptive wiiters, and is seen in Ampliitetras, 

 Biddulphia, Podosira, and Melosira. In the last two genera. Prof. Smith 

 tells us, the persistence of this cii^cular band is '' eminently conspicuous, 

 retaining the frustules after self- division in a geminate union until the self- 

 dividing process is renewed." 



Contents of Feustules. — Nucleus, supposed Digestive Sacs, Heproductive 

 Vesicles, ^r. — The organic membrane of the frustules of Diatomeae, strength- 

 ened externally by the silicious plates, encloses within its cell-like cavity a 

 soft mucilaginous substance filled with numerous granules and globules, and 

 usually of a yeUow-brown or orange-brown colour, but at times of a 

 green hue, and technically known as the * endoclirome,^ or in Klitzing's 

 phi'aseology, the ' gonimic substance.^ The granular matter is particu- 

 larly aggregated about the organic waU, leaving the central portion more 

 clear. In this clear central space is a transparent vesicle, representing the 

 nucleus of the cell, having the granules frequently collected around it in an 

 annular form. Nageli states that the nucleus, enclosing a nucleolus, lies 

 sometimes free in the centre of the frustular cavity, but at other times is 

 affixed at one spot to the wall, and therefore ' parietal.' He also describes 

 two sorts of nuclei, viz. primary and secondary, attributing to the former the 

 active part. Schleiden represents the nucleus to be primarily concerned in 

 the original formation of the cell, as well as in its subsequent multiplication 

 by seK- division. 



Among other elements of the endochrome are more or fewer rather 

 translucent globules, which Prof. Smith believes, like Kiitzing, to be secre- 

 tions of the cell, of a fatty or oily composition, and to be the source of the 

 peculiar odoiu* emitted on burning the Diatomeae. In support of this 

 \'iew Kiitzing states that he has occasionally seen two coalesce, proving the 

 absence of proper walls, and expresses his conviction that these corpuscles 

 are akin to the amylaceous secretions of the Desmidieae and Confeiwae and the 

 starch-granules of the higher vegetables. 



These globules are smaller than the nuclear space, and occupy a pretty 

 constant and definite position. " The number of these globules is frequently 

 four, often placed near the extremities, or more rarely clustered round the 

 central vesicle." Meneghrni (op. cit. R. S. p. 364), alluding to these vesicles, 

 states them to vary in number, size, and disposition at different stages, and 

 according to various conditions, even under the eye of the observer. 



These apparent oU-globules were called by Ehrenberg male sexual glands 

 or testes, whilst those other vesicles distributed within the mucilaginous 

 matter, often about the nucleus, were named stomachs. The latter idea he 

 based especially on a series of experiments to introduce colouilng matter 

 into the interior of the frustules, in which he believed he succeeded. The 

 species mentioned are Navicid a gracilis, N. Ampjhishoena, N. viridida, N.fidva, 



