ZOOLOGY AND BOTANY, MICROSCOPY, ETC. 951 



nearly as rectangles, and the latter giving more of the spindle shape. 

 This is analogous to the difference noted in other genera, the outer 

 view of hexagonal markings being usually nearly circular, whilst the 

 inner shows the angles more clearly. In Epithemia iurgida, as 

 found in Moller's preparation from the Sodertelge mud, the frame- 

 work of the shell is a nearly rectangular lattice, the areolae showing 

 all the peculiarities of light which have been described in Navicula 

 lyra, and the fracture often shows the ends of the framework 

 sticking plainly out beyond the sides of the adjacent dots. The same 

 may be seen in the elongated areolae of Amphora ovalis of the larger 

 varieties. In Gucconeis sptlendiduni the hexagons are as distinctly 

 formed as in Coscinodiscus, and in C scutellum the areolation varies 

 from coarse to fine with the diminishing size of the valves, giving a 

 series analogous to those which arc found in G. subtilis, and one in 

 which the fracture is as plainly through the dots, whilst the evidence 

 of relative thickness or thinness of the silex from the colour is as wo 

 have found it in other cases. 



But to complete the list of species in which I have found the 

 tests of fracture and of colour supporting the theory of areolation of 

 the diatom shell, and contradicting that of solid spherules, would be 

 too much like making a catalogue of all in which the details are largo 

 enough to give a well-defined outline to a broken edge. In the 

 progressive series of fine markings we sooner or later reach the 

 point where the thinness of a film causes it to be lost in the general 

 background of the field, or where the prismatic edge of a fracture 

 makes diffraction enough to fringe it with lines of colour or of ap- 

 parent shadow, which make every cautious observer hesitate to affirm 

 whether the boundary is in or beyond one of the striaD. The fringes 

 move with the slightest motion of the fine adjustment, and the inter- 

 pretation of what we see is more or less modified by the preconceived 

 theories of the observer. I have intended to draw my examjDles of 

 facts from specimens found clearly within this limit of doubtful 

 discrimination. I am myself satisfied that in the coarser specimens 

 of difierent species of Plturosigma careful illumination and accurate 

 adjustment of good lenses show the same characteristics of structure 

 at broken edges of shells which I have described in the larger and 

 bolder forms. In regard to this, however, I admit there is room for 

 disjiute. In the matter of the colour test, on the other hand, the 

 evidence seems to me clear. If the objective is well adjusted, and the 

 median line is brought into focus, so that it appears a greenish white 

 line of nearly the same tint as the general field, the dots which aro 

 near enough to it to be in the same plane aro found to have the same 

 colour. In P. formosum, P. halticum, P. attenuatum, and the varieties 

 closely allied to each, the reticulation seems to be thickened upon 

 the outer edges of the lines, so as to leave a cup-like depression in tlie 

 interstices, which is yet consistent with double laminae below. Wo 

 have seen that in Eupodiscus this thickening becomes so great as to 

 be quite opaque. In Aulacodiscus oregonianus, and in A. orientalis, 

 it is Komotinios found thick enough to give a decidedly dark colour 

 to the reticulation (jf the surface. In media of higher refractive 



