ZOOLOaY AND BOTANY, MICROSCOPY, ETC. 949 



wliicli is part of the firm silex on one side and laj)s under the other 

 side in a way simiLir to the ' rabbit ' in joinery. 



Of the dotted Naviculce, N. lijra may fairly be taken as the type. 

 Its beautiful regularity of form, and the clearness and boldness of its 

 marking make it a very profitable^subject for careful examination. It 

 is easy to get somewhat varied appearances by ditferent uses of the 

 light and changes of focus of the objective, but if we use the narrow 

 central pencil of light and care in focusing, its characteristics will be 

 found uniform and unmistakable. Its lyrate hyaline figure in the 

 middle of the valve takes the pink tint. The dots are found to 

 be between costfe which are fully as wide as the dotted interspace, and 

 these have the same colour as the lyrate figure. Find a broken shell 

 and focus carefully upon the broken margin. Oftentimes the costae 

 will be found to project beyond the interspace, showing its greater 

 strength, and confirming the evidence to this efiiect which is found in 

 its deeper colour. When the focusing gives us the costfe as well- 

 defined ribs of even width, and a broken edge is also most sharply 

 defined, the dotted intersj)ace will aj)proximate to a ladder-like ap- 

 j)earance, the dots having a sub-rectangular form, and being separated 

 from each other by septa considerably narrower than the costfe 

 between which they lie. The term ' sub-rectangular ' which I have 

 used must not be taken too literally, for the figure of the dots is that 

 of a circle somewhat flattened on four sides. Assuming that the 

 median line is a groove in the valve, and focusing upon it so that the 

 light coming through it shall correspond nearly to the general field, it 

 will then be found that the dots nearest this line and most perfectly 

 in the same plane show the same colour, an item of evidence that they, 

 too, are thin places in the shell. But the line of fracture gives still 

 stronger proof. I have before me a broken valve of N. lyra, in which 

 a segment is entirely gone, bounded by the median line for, say, half 

 the distance from the end of the shell to the central nodule. Then 

 the broken margin runs irregularly off to the rim of the shell. On 

 the other side a wide crack extends diagonally from the median line a 

 short distance, then runs straight out to the rim. This crack (exa- 

 mined with a 1/15 objective) zigzags through the dots in the first part 

 of its course, and in the straight part runs indisputably through the 

 dots and between the straiglit costfe. The broken edge of the other 

 side of the shell shows with equal clearness that the fracture is through 

 the dots. I have many such cases noted, with great varieties of frac- 

 ture but all indicating the same fact in regard to structure, viz. that 

 the dots arc the thin and weak places in the valve. 



Another point to be noted is that whilst the radiant costal of iV. 

 lyra arc straight, making also straight transverse striation, when 

 viewed with a low power, the longitudinal septa between the dots are 

 not regularly continuous; consequently, when light is thrown trans- 

 versely across the shell a low power shows longitudinal striae, but 

 wavy instead of straight. This is also the case with the striation of 

 N. firma, N. cusj^idata, N. rhomhoidcs, and Frustulia saxonica when 

 examined witli higli powers, and with the Nitzscliias of the form of 

 N. scalaris, N. linearis, &c., of which the coarser specimens show 



