38 OF THE CHARACTERISTICS OF THE DIATOMACEiE. 



long cylindrical surface, «, b, tlie ' front view,' * and the 

 base, c, b, the l side view,'j* we see that the two aspects 

 bear the same relation, as do the long and short axes of the 

 cylinder, to each other. 



Now, if we suppose the cylinder to be divided into 

 several very thin sections, a d, d e, e b, &c, in lines per- 

 pendicular to its longer axis, it is clear that, so long as the 



Fig. 45. 



€ 



rows of sections remain attached to each other, the cylin- 

 drical surface, a, b y appears to the eye of the observer the 

 largest, and is in fact the 6 front.' But the whole aspect of the 

 object is changed as soon as the segments are separated 

 and placed apart, as f, g. Then the base exposes the 

 larger surface, /z, and is really the ' front ' to the observer. 

 However, to avoid confusion, and to have fixed ideas in 

 reference to these parts, the base (or part through which the 

 knife passed in our imaginary section) is held to be the 

 1 side,' while the name of ' front ' is appropriated to the 

 part which is cut through, notwithstanding that the so- 

 called i side ' may be more conspicuous and more frequently 

 exposed to view. As the Diatomacese propagate them- 

 selves by self-division, thereby forming longer or shorter 

 rows of frustules, the diagram given above affords a fair 

 representation of the manner in which the frustules are 

 attached by their i sides,' as the collector may easily satisfy 



* This is the 'primary' side of Kiitzing and the 'secondary' of 

 Rabenhorst. 



t This is the ' primary ' side of Rabenhorst, and the ' secondary ' 

 of Kiitzing ; so widely do authors differ in their nomenclature. 



