46 GEKEEAL HISTOEY OF THE INFUSORIA. 



to the age and to the size of the valves, and both size and figure are consi- 

 derably aifected by circumstances of growth, of locality, and the like. A 

 writer in the Microscoj^ical Journal (1855, p. 307) suggests that the number 

 of striae on the entire valve may supply a more stable character ; yet even 

 on this point we are wanting in direct observ'ation to show that this number 

 may not be affected by accidental circumstances. 



Although " the relative distance of the strise and their greater or less dis- 

 tinctness " be accounted by Prof. Smith of specific importance, yet he is 

 obliged to admit (J. M. S. 1853, p. 133) that it is by no means certain that 

 these features may not to a slight extent be modified by localities and age, 

 and is disposed to believe that they are certain guides only when we have 

 made allowance for these conditions, and that while they are constant in 

 fnistules originally from the same embryo, they may slightly vary in those 

 which owe their birth to difierent embryonic cells. It is also worthy of 

 note that, in certain instances, e. g. in Odontidium hyemalis, Ejnthemia A^xus, 

 &c., the costse are frequently more numerous on one valve than on the other. 



Other illustrations of the variation in the number and in the distinctness 

 of striae in the same species are to be found in the late lamented Dr. Gregory's 

 valuable papers (T. M. S. 1855, p. 10). The relative position of striae — if 

 parallel or radiate, their monihform or confluent character, their equal and 

 general diffusion over the entire surface, or theii' absence at parts, are other 

 circumstances available for the pmposes of classification. 



Besides striation, the other descriptions of superficial markings are resorted 

 to for specific and generic characters. Such are the presence or absence of 

 a median band, with central and terminal nodules, the existence of a trans- 

 verse band, the figure, relative position and aggregation of the areolae or cells 

 of the surface. The median and transverse bands have been employed by 

 aU systematists, and would seem well suited to furnish characteristics by their 

 constancy. The same may be said of the pore-like spots or nodules. Kiitzing 

 went so far as to make the presence or absence of a central nodule or um- 

 bihcus the turning-point in his grand division of the Diatomeae into um- 

 bilicated and non-umbilicated. 



Lastly, speaking generally, the precaution intimated by Prof. Walker- Arnott 

 (p. 39), of having specimens, intended to be compared together for the determi- 

 nation of specific forms, similaiiy prepared, must ever be borne in mind where 

 the superficial markings are referred to for characters; otherwise an excessive 

 and erroneous multiplication of species, and a deplorable confusion will result. 



We have already seen that the connecting membrane is not an essential 

 segment of a Diatomaceous frustule, but an after- development in connexion 

 with the process of self- division; yet, notwdthstanding, it is so frequently 

 present, and in many examples its dimensions and characters are so marked, 

 that it supplies an important element in specific and generic descriptions. 



In the circular and discoid Diatomeae, it assumes the form of a continuous 

 ring (XI. 40, 42), but in many oblong and navicular frustules it is itself 

 oblong or navicular, having a figure the reverse of the valves it is placed 

 between (XII. 17, 24, 31 ; XIII. 5, 6, 7). In these latter and in other 

 instances it is frequently separable into two portions, at the opposite extre- 

 mities of the frustules where the silex is absent ; and hence it is that the 

 shells of the Diatomeae have been described by Meneghini and other writers 

 as composed of foiu' segments. 



In general, the proportion of silex in its constitution is less than that in 

 the valves; and the existence of markings— of areolae, striae, and the like— 

 is also much rarer. Where they do occiu-, they furnish useful particulars 

 in defining species and genera. The small development of the connecting 



