FROM THE WEST COAST OF AFRICA. 151 



the primary ' rays,' it will be found that the same granules also 

 radiate from the centre to the margin, bi-furcating at different 

 distances, in accordance with the varying forms of the valves. . . . 

 The decussating aspect of the lines of granules depends in great 

 measure on the forms of the valves ; if the processes be on a raised 

 or mammate surface the lines will be more curved or arched than 

 when they rise from a flat surface, and in such case the decussation 

 is more striking to the eye." This arrangement of radiant, oblique, 

 and decussating lines of granules is common to A. Petersii, 

 A. Kittoni (though in some varieties very much more than others, 

 and most striking where the processes are most numerous), 

 A. Sollittianus, and in the new diatom from West Africa. The 

 chief distinction, as it appears to me, between undoubted specimens 

 of A. Kittoni and the new form is in the processes, which in the 

 first are (as Mr. Hardman describes them) like " eye-brows," and 

 the last are distinct " hoods." 



I will now give Professor Walker-Arnott's description of A. 

 Johnsonii — "Disc pale with a circular perforation-like umbilicus, 

 and ci'owded radiating series of granules, becoming more numerous 

 as they proceed outward so as to appear forked ; processes within 

 the margin, roundish, small. Algoa Bay guano. The rays near 

 the margin become more numerous with smaller granules, so as to 

 look like stria? ; sometimes the processes appear within a faint 

 circle. A. Johnsonii somewhat resembles A. Kittoni, but is less 

 hyaline, with more conspicuous granules, and processes more dis- 

 tant from the margin." Figure 7 has been drawn from Mr. Cole's 

 specimen of this species. 



All the specimens of A . Johnsonii from Algoa Bay guano are 

 small, none that I have seen are very much larger than the speci- 

 men figured. The valve appears to be without any raised portions, 

 for there is no appearance of oblique or decussating lines or curves, 

 and its distinct border nearly one-sixth of its whole diameter of 

 very small granules is a feature that distinguishes it from A. 

 Kittoni, as also does its umbilicus. The processes are covered by a 

 " hood," which is so hyaline that it is easily overlooked, and so 

 brittle that it is my impression that I have not seen a single speci- 

 men in which they were all perfect. At " a " in figure 7, I have 

 represented the appearance of a process that has bee/t broken off. 

 From his description Dr. Arnott seems to have misunderstood their 

 character, for he describes them as " roundish, small," and then 



N 



