72 p. T. CLEVE, SYNOPSIS OF THE NAVICULOID DIATOMS. 



where is a large foramen. The foramina form in the large rspecies, a more or less distinct or 

 broad longitudinal band across the strise. 



The central nodule consists of a conical silicious mass, projecting in the inside of the cell, 

 usually excentric, so that one margin of its base (the median margin') coincides with the axis of 

 the valve. Between the nodules of both valves, in the interior of the cell, is the median plasma- 

 mass, containing in its middle the nucleus. According to Otto MOller the central nodule has, on 

 the top of the cone or its median side, an open shallow furrow, in both ends of which is a pore, 

 by which the plasma communicates with a system of closed channels in the central nodule and, 

 by means of similar channels in the median line, with open fissures in the parts of the median 

 line which are between the centre and the ends. From these pores pass vertically, through the 

 solid mass of the nodule, two channels (the channels of the central nodule) which terminate on the 

 exterior of the valve as two bulb-shaped pores (the median pores). The channels of the central 

 nodule each give rise to two parallel, closed, channels, one above the other (the exterior and in- 

 terior channels). 



Two similar channels also open into the terminal nodules. The exterior of these channels 

 is at its end bent towards that side of the valve, on which the central nodule is situated, and 

 terminates in an oblique, sometimes spirally twisted, fissure (the terminal fissure), which bends 

 round the terminal nodule. The terminal fissures are in most cases turned to the same side, but 

 in some few cases in contrary directions. Their shape is somewhat different in different groups, 

 sometimes curved and comma-like, sometimes straight, giving to the ends of the median line a 

 bayonet-like shape. 



On the lower side of the curved end, or opening of the exterior channel, opposite to the 

 terminal fissure, is a triangular fissure (the funnel) which passes obliquely through the terminal 

 nodule, and opens with its pore-like apex into the plasma-mass at the ends of the cell. The 

 median part of the funnel communicates with the interior channel of the terminal nodule. 



The exterior and interior channels from the central to the terminal nodules are connected 

 by a fissure of more or less complicated structure. In many cases this fissure forms a filiform line 

 and seems then to be simply a vertical fissure. In other cases it is broad, and lies in an oblique 

 position. In some of the larger species the fissure is of a more complex nature, being formed by 

 the junction of short knife-like lamina; projecting from both halves of the valve. From the 

 half of the valve, on which the central nodule is situated project three, in some cases two, such 

 laminae, alternating with two (or one) laminae from the other half of the valve. In some of the 

 larger species the anterior laminae on the half of the valve which contains the central nodule 

 cover the laminae on the other haK of the valve on two portions of the median line, thus giving 

 rise to a biundulated median line. 



The genus Pinnularia comprises a great number of forms, both marine and fresh-water, the 

 latter being very predominant, forming an important and very characteristic part of the fresh- 

 water diatomaceous flora, both recent and fossil. The marine forms are not all closely connected. 

 One group among them is characterized by the peculiar form of the large terminal fissures, pro- 

 jecting from the ends of the median line at an angle of about 90°. To this group belong P. Tre- 

 velyana, P. rectangulata, P. groenlandica, and in a less degree P. cruciformis. The latter seems to 

 be connected with P. quadratarea, having indistinct terminal fissures. Smaller marine forms bear 

 a great resemblance to the smallest fresh-water Pinnulariae of the group Parallelistriatce. Among 

 the marine forms are a few more or less panduriform, which appear in several respects to be 

 akin to panduriform species of the genus Caloneis. 



The fresh-water forms pass into one another to a great extent, so that the definition of good 

 or distinct species or groups is a matter of the greatest difiiculty or almost impossible. Still, I 

 think some groups of forms may be distinguished, although closely connected with each other. 

 Such groups are the following: 



