Structural Variation among the Biffiugian Rhizopods, 233 



same time that one of the angles is truncated obliquely, thirdly, 

 plates approaching in outline rectangular prisms ; and lastly, 

 bodies in which the tendency to crystalline outline is lost and 

 they pass into oblong colloid discs somewhat depressed at their 

 centre. 



We now come to the forms shown at figs. 28 & 29, in which 

 the surface of the tests is studded with the oblong bodies, some 

 of which, however (as is more manifest in the larger specimen), 

 exhibit a faint but nevertheless definite approach to elongated 

 hexagonal prisms*. These measure about -.^^'outh of an inch in 

 length. They are unmixed with any other bodies, and arranged 

 side by side with a certain degree of regularity. In these the 

 central depression is also distinct. The larger specimens (fig. 29) 

 are from Greenland. The smaller, in which the discs are not 

 more than about -j-oVoth of an inch in length, are from Hamp- 

 stead. 



In fig. 33, which also represents a tolerably frequent variety 

 of test, the discs attain their maximum of regularity as regards 

 shape : all are circular, or very nearly so, and exhibit the cen- 

 tral concavity in a very marked manner ; indeed they resemble 

 blood discs in several particulars, but of course the resemblance 

 is merely apparent. They are of varying sizes, the largest 

 averaging about -o-aVo^h, whilst the smallest are not more than 

 1 ooou ^^t of an inch in diameter, the larger ones being generally 

 surrounded by groups of the smaller, although, as seen in a 

 number of specimens, there is evidently no particular order in 

 which they are arranged, beyond that resulting from their taking 

 up positions side by side, and all, without exception, resting on 

 their flat surfaces. 



But it is in the series of forms already referred to as built up 

 in a great measure of minute diatoms that we find the clue to the 

 origin of the colloid discs and rectangular plates. In figure 32 

 the diatoms, as will be seen, are interspersed amongst the circular 

 discs. The very important fact reveals itself, howevei*, that 

 some of the diatom-valves are becoming gradually metamor- 

 phosed, that is to say, exhibiting a gradual passage from the 

 typical outline of the little Eunotia to one more closely ap- 

 proaching an irregular cylinder, the cylinder then passing into 

 the elongated disc, which is at times distinctly hexagonal, and 

 finally the elongated disc passing into the circular one. The 

 specimen figured is not so calculated to show the transitionary as 

 the unmetamorphosed state of the diatoms ; but in some indivi- 

 duals, and more especially in crushed specimens of these tests, 

 every stage of transition may be clearly distinguished. As 



* It is not improbable that the tendency to assume this hexagonal form 

 may result from pressure of the discs one upon the other. 

 Ann. l^ Mag. N. Hist. Scr.3. Fo/. xiii. 16 



