IN VEGETABLE BIOLOGY. 61 I 



secondary rings (fig. 37, a). The secondary rings, although 

 usually circumferential in position, are not always so. Eig. 41 

 shows a small membrane, near the centre of which are placed 

 what appear to be small secondary rings ; a string of proto- 

 plasm is attached to one of them. 



The larger of the cells surrounding the rows of wide ones 

 communicate with the latter, and with their fellows, by means of 

 rings which in some cases at least are compound (fig. 41). The 

 rings between the cells with narrow lumen are small and simple. 

 The small epidermal and immediately underlying cells are con- 

 nected by fine nodule-bearing threads similarly to those of Cera- 

 mium ruhrum (fig. 27). 



Working with dried material, I have laboured under a distinct 

 disadvantage in studying the development of these rings. All 

 that I have learnt about the matter is this : — Below the almost 

 hemispherical apical cell of a growing branch are two demilune 

 cells, one beneath the other. Beneath this is a similar but larger 

 demilune cell which divides into five cells, two on either side of a 

 central scutcheon-shaped one. The outermost cell on each side 

 takes part in the formation of the thin lamina of the frond; the 

 central cells rapidly elongate in the direction of the axis of 

 growth, and after a time a transverse septum divides each of 

 them in turn (fig. 42). Although the cells for some time after 

 their formation are in very close apposition, it is possible, with 

 the aid of chloriodide of zinc and by careful focusing, to see nume- 

 rous small dots of protoplasm between the newly formed cells. 

 These ultimately disappear with the exception of a single fine 

 thread, at the centre of which is swung a small bright nodule of 

 protoplasm (fig. 42). The intercellular threads appear to persist 

 until division of the cells is definitely completed, whereupon, 

 except for the one thread, they rapidly vanish ; by this means con- 

 tinuity is maintained throughout the whole thallus. Until now 

 cell-division has been at right angles to the plane of the thallus ; 

 but by a series of divisions parallel to that plane the formation 

 of a midrib is announced. In the earliest condition of distinct 

 continuity between the largest as also between the other cells, I 

 have never seen more than one strand of protoplasm. Some little 

 way below the growing-point the connection between contiguous 

 large cells is as shown at fig. 43, where at least three fine strands 

 of protoplasm pass each of them through a small nodule appa- 

 rently without undergoing division. At this stage the ring is 



