COSCINODISCUS. 527 



lines of the wedge-shaped fascicles are lengthened till those of adjacent 

 fascicles touch. The small spine is at the marginal end. of the middle line of 

 the fascicle. Bevelled rim beyond the spines similar to last. No pseudo- 

 nodule. 



?. Coscinodiscus radiolatus Ehr. Marking of the disc similar to 

 the last, except that the fascicles are each formed by a radial line, with parallel 

 lines on one side of it only, the fascicles being all similar to each other and 

 symmetrical. The small spine is at the end of the radial line forming the side 

 of the fascicle. 



4. Coscinodiscus lineatus Ehr. The disc covered with alveoli in 

 straight lines parallel to a diameter and arranged in quincunx, so that where 

 the round alveolus becomes hexagonal the surface of the disc resembles a 

 true honey-comb, the most prominent lines in appearance being parallel to 

 the given diameter ; a loose circle of small spines marginal or intra-marginal. 



5. Coscinodiscus radiatus Ehr. Marking of the disc made up of 

 hexagonal alveoli in radial lines, which, starting from a central rosette, bifurcate 

 as they go outward, so that there are rarely continuous straight radial lines of 

 alveoli. In robust specimens the upper surface of each alveolus has a dotted 

 appearance, caused by finer secondary pits or marks. 



6. Coscinodiscus centralis Ehr. Marking of the disc of hexagonal, 

 round or subquadrate alveoli, which, starting near the centre, go outward in 

 continuous radial lines, new radial lines being intercalated as room is made 

 by the outward divergence. Secondary maiking in robust specimens: in 

 large examples the central zone is thin, and the outer zone relatively robust 

 and thick. 



7. Coscinodiscus marginatus Ehr. Marking of the disc sub- 

 hexagonal, alveoli without traceable scheme of radiation, but in largest 

 specimens approaching Cos. radiatus. Margin strongly recurved and produced, 

 so as to make a more and more deeply cup-shaped valve in each new frustule 

 formed by fission, the small valves being thus most deeply cupped in a given 

 brood, and the large valves most flat. 



I have already several times indicated my way of looking at the matter, 

 and I believe that whenever an easily recognised division can be formed 

 whether called tenus or subgenus or section it is useful to form it in order 

 to restrict the field of research, and to render easier determinations. Thus 



