36 GENEKAL HISTORY OF THE INFUSORIA. 



The same authority has the following remark on the process of ramification : 

 he says (i. 75), " When self-division {i. e. of the frustules) is completed, the 

 extension of the filament below the fnistules is suspended, a joint or arti- 

 culation is formed at the base of the dividing frustule, and each of the half- 

 new frustules begins anew, in its progress towards special self- division, the 

 secretion of a new joint or internode ; and a dichotomy is the result." 



The occurrence of the double condition of union of frustules in a con- 

 catenate manner and of attachment by a pedicle is illustrated in the genera 

 Achminthes (X. 201, 202), Striatella (X. 203, 204), BhabcJonema (XIII. 27), 

 and Podosira (II. 45). In Melosira also, attached species occur ; and Prof. 

 Smith inclines to the opinion that all filamentous Diatomese are stipitate on 

 their first production. In the second stalked genus cited, viz. Striatella, the 

 stalk attains the highest development, but remains slender and unbranched. 

 Between this most developed foim and the mere nodules of attachment in the 

 genera Achnanthes and Melosira, every intermediate phase is encountered. 

 In any one species, however, there is no positive determinate length of the 

 stipes, for this varies according to the idiosyncrasy, vigour, and external con- 

 ditions affecting the organism; consequently characters derived from the 

 dimensions of the stems can have no specific value. 



There is a large section of Diatomeae in which the frustules are diffused 

 throughout a mucous or muco- gelatinous mass, rarely confusedly, but mostly 

 in a definite manner, usually in thread- or tuberlike branches, which nor- 

 mally ramify in a dichotomous fashion, and resemble on a minute scale the 

 tufts formed by many large sea-weeds. This peculiar aggregation is the 

 consequence of the large production and subsequent persistence of the mucus 

 which is thro^Ti out when the system of reproduction, whether by sporangia 

 or by fission, takes place. Histologically, therefore, it is homologous with 

 the pedicles and connecting nodules or isthmi thrown out duilng the act of 

 seK-division, as also with the mucous stratum, which still very often persists 

 when that act is complete, around specimens of Cocconeis, Chcetoceros, Melo- 

 sira, Fragilaria, Striatella, &c. 



The tissue thus composed of mucus and enclosed frustules constitutes what 

 is called (from analogy with the large Algae and other Crj^togamic plants) 

 the frond, and affects various shapes, in some measure characteristic of the 

 genera. Thus, in one of those so-called frondose Diatomeae, viz. DicMeia 

 (XY. 30, 31), it is membranous and leaf-like, and resembles a species of 

 IJlva; in Mastogloia, filiform with nipple-like expansions; in Encyonema 

 (XIV. 22), Homceocladia (XIY. 37, 38, 47, 49), and Sehizonema, filamentous 

 and more or less branched ; in Colletonema suhcohcerens, globose. Again, when 

 filamentous, the ramifications differ much in thickness and in expansion, and 

 in the extent of adhesion between the branches ; where these are long and 

 slender they are called ' capillary,' and where contiguous branches coalesce, 

 they give rise to a submembranous condition. The degree and mode of 

 division, the collection of the branches into bundles (i. e. fasciculi), or, on the 

 contrary, their loose or diffuse arrangement, supply useful characters in the 

 distinction of species. Again, the fronds differ in consistence, being in some 

 genem or species more rigid, setaceous, or robust, in others softer, flaccid, 

 and more delicate : these opposite conditions furnish Prof. Smith with grounds 

 for the division of the genus Sehizonema into two tribes. 



The disposition of the fi^ustules within the mucous investment supplies 

 other important distinctions. Thus in DicMeia it is irregular ; in Mastogloia 

 each little frustule occupies '' the summit of a little nipple-Hke cushion of 

 gelatine ; " in BerTceleya (XIV. 34, 35) the frustules are densely packed in 

 the filaments ; in Encyonema they occur mostly " in single file, except 



