W. M. BALE ON SOME OF THE DISCOID DIATOMS. 35 



dividing it into zones. Sometimes the markings are denser on 

 one of these zones than elsewhere. Very often the zones form 

 hyaline bands on which the granules are wanting, and the 

 structure may be further complicated by the addition of radial 

 hyaline bands, e.g. two hyaline zones may be joined by a number 

 of equidistant radial hyaline areas so that the space between 

 them is divided into a circular series of sub-rectangular com- 

 partments ; or a broad circular zone may be filled with hyaline 

 patches of all sorts of irregular shapes. The radial series of 

 granules may be all curved in a spiral fashion (a variation 

 which also occurs in C. Ehrenbergii), and I have specimens in 

 which the central portion, as far as the first circular interruption, 

 has the moniliform series all contorted in the most extraordinary 

 manner. As in A. rex, etc., the subulate areas may be either 

 darker or lighter than the rest of the valve. There may be a 

 small central area, or the whole centre of the valve may be 

 sparsely and irregularly marked. 



I find that in some slides concave and convex valves are mixed 

 about equally, leading to the conclusion that the two forms 

 represent opposite valves, as in A. rex, but in other gatherings 

 I find many concave valves to every convex one. Rattray de- 

 scribes the valves as flat in the centre and otherwise convex, but in 

 numerous cases the convexity (or concavity) is uniform throughout. 

 Asteromphahis. In this genus the lines which radiate from 

 about the head of the centro-lateral area to the apices of the 

 areolate compartments have been assigned too much value in 

 classification. Whether they originate from a single point, or 

 whether they bifurcate, is absolutely immaterial, and the presence 

 of geniculate bends in their course is, in some species at least, 

 equally unimportant. A. Hookeri, which is not rare in one of the 

 "Challenger" Antarctic soundings, illustrates this. The forms 

 with six, seven, eight and nine rays, which represent four of 

 Ehrenberg's "species," also a ten-rayed form, occur in slides 

 which I have prepared from this material, and 1 find the 

 geniculations of the radial lines very marked in some valves, 

 while others show no trace of them ; others again exhibit a mixed 

 condition. A good deal seems to depend on the size of the valve, 

 the geniculate lines being most common in the smaller ones. 



Certain species are subject to variation in the outline. 

 A. Cleveanus, as figured by Schmidt, has a rather narrow ovate 



