(368 Transactions of the Society . 



wails several results were obtained witii the sections. These 

 two new disks have the same sculpture as the two of the outer 

 valves, except that all outlines are much finer, and the chambers 

 are also not exactly of the same definite height. One observes that 

 the chamber-walls in the two young disks sometimes touch each 

 other, sometimes not. The manner in which all these parts are 

 connected with one another is very peculiar. I must here refer 

 to what I say further on about Achnanthes. With Coscinodiscus 

 there is no doubt that at least the one of the two newly formed 

 cells is completely closed outwards. The basal membrane of the 

 one outer valve goes continuously in equidistant curves into the 

 basal membrane of the corresponding inner valve ; at the bend 

 a few delicate lines are seen which are the sections of the walls of 

 the considerably reduced marginal chambers. Out of this portion 

 the new girdle-band must be developed, as hitherto it has not been 

 detected. One observes in the other newly formed cell (the lower 

 one in the fig.) existing in all central sections a break, which the 

 membrane of the new valve makes at the bend, and this gives the 

 appearance as if the basal membrane of the old valve was con- 

 tinued here directly in the inclosed girdle-band, which is probably 

 the case; but one also observes here a break. Nevertheless I 

 believe that this second cell is closed at the bend of the membrane, 

 a conviction to be gained only by very careful study of the entire 

 section series. Future investigators who will work with patience 

 have an inexhaustible field before them ; they will have to turn 

 their attention specially to the gradual development of the chamber- 

 walls and determine whether, for example, as Mtiller indicates for 

 Triceratium, network arises from a basal membrane at first 

 smooth, or whether it is more probable that at first hollows are 

 developed in the membrane which afterwards become chambers 

 open at the top. It is interesting to see in this fission specimen 

 how, during the process of division, the lumen of both cells was 

 reduced almost to nothing, and very nearly the entire inner space 

 of the former mother-cell was filled up with the substance of the 

 two new walls. A coalescing of the two disks during the imbed- 

 ding in gum cannot be entertained, inasmuch as they would show 

 cracks at the edge, as old valves are unusually brittle. If we com- 

 pare with this an ordinary undivided specimen, from one of which 

 I have given a nearly median section out of the marginal portion, 

 fig. 24, the difierence is at once apparent. In other respects such 

 a section displays very little variation from the above representa- 

 tion. The first section through the same specimen, which shows 

 the girdle-band of the surface and the minute chambers in the 

 convex edge, is given in fig. 25. I could not obtain collodion 

 casts of any service, nor could I succeed in fixing permanently 

 colouring matters in the spaces of the chambers. If the latter 



