Dr. Walllcli on tJie Rhizopods. 461 



this case and also in that of the single Cocconeis valve, which^ 

 however, still exhibits the median line and its nodules. 



In plate xxii., representing Nehela coUaris^^gs. 13, 18, 19, 

 and 20 (the two last "magnified 850 diameters '"'), are to be 

 seen pertectlj rectangular rods (called " YineRV plates" in the 

 descriptive note), together with oval and round plates of 

 various sizes, the lai'ger ones exhibiting the curves of oval and 

 circular diatom-valves. The two highly magnified figures (19 

 and 20) show the perfect angles of the rectangular plates and 

 the tabular surface of the ovate and circular ones. B at the most 

 important piece of evidence is furnished bj the test (fig. 18), 

 described as " shell of narrow rectangular and oval plates^ from 

 which a broad strip was broken away, showing that the fracture 

 follows the intervals of the plates.'''' It must be remembered 

 that the test is a large one, and the fractured aperture extends 

 diagonally nearly from its apex to its base, not a single plate 

 appearing to be broken across, but the irregular outline con- 

 sisting of zigzags, each of which answers p)''^o tanto to a side 

 or end view of the plates. Hence we may reasonably assume 

 that the whole of these plates are identical with the plates of 

 Diffiugia symmetrica in being perfectly rigid, strong enough 

 not to yield to considerable force, and therefore necessarily 

 composed of a large percentage of siliceous or other mineral 

 matter in combination with the colloidal basis of the test. 



But the nearest approach to the perfectly square figure of 

 the plates in Diffugia symmetrica occurs in a specimen of 

 Nehela flalellulum^ represented in plate xxiii. fig. 19, a broad 

 balloon-shaped variety, on the front of which there are eight 

 perfectly-formed '"'' square plates " scattered irregularly among 

 a crowd of " circular, oval, and linear plates." In size and 

 appearance the square plates exactly resemble those of D. 

 symmetrica. Prof. Leidy (at p. 153) remarks : — " Occasion- 

 ally I have found specimens in which quadrate plates^ like 

 those of Quadrula symmetrica^ were mingled with the more 

 usual structural elements as seen in fig. 19," i. e. in the figure 

 I have been referring to. Nevertheless, we know Prof. Leidy 

 has in his definition of ^^ Quadrula " declared that the test is 

 " composed of thin square plates of chitinoid membrane^'' 

 apparently forgetful of the following very circumstantial 

 statement made by him in reference to N^bela collaris, the 

 very form in which ' the quadrate plates, like those of Quad- 

 rida symmetrica^ occasionally occur." 



" In breaking the shell," he observes, " the line of rupture 

 follows the outlines or intervals of the disks and plates. The 

 shell [of Nehela collaris] appears to he silicious, as it remains 

 unchanged when exposed to the action of heated sulphuric and 



