STTJDTJ OF THE GENICULA OF CORALLIN.i:. 29 



addition of the secondary lamella. The middle lamella follows the 

 growth, though it is much thinner than in the articular cells. 



The growth of the cells seems to be of such great rapidity that 



the cortical cells around the geniculum can no longer bear the strain. 

 The result is t lie cracking of the cortex. 



Regeneration seems to take place at the overhanging margin 

 of the circumgenicular cortex after the splitting has occured. On the 

 freed ends of the cortex we actually observe the remnants of the decav- 

 ed cells which have broken during the formation of the fissure ; and 

 generally we find the cells in the circumgenicular cortex dividing in 

 the anticlinal and periclinal direction (figg. 4. 6 and 10). This 

 circumstance, no doubt, effects the "linear" genicula, exposing a 

 comparatively smaller portion outside, the larger part of the length 

 being hidden under the circumgenicular cortex. 



The best developed example of the circumgenicular cortex is 

 found in Amp. stelligera. In this species we find a highly thickened 

 portion at each distal end of an articulus. This thickening is due to 

 the unusual increase in the number of the cells of the circumgenicular 

 cortex, (fig. 8). 



While the geniculum is yet very young or the development is 

 not comparatively rapid, the cortical layers remain around the 

 geniculum (fig. 5). !) These layers in most of the species are destined 

 to drop off sooner or later. Harvey 2) describes a plant with calcareous 

 granules upon the genicula. His plant had probably the circum- 

 genicular cortex split into several pieces, remaining attached upon 

 the surface of the genicula. Very often we find similar cases in 

 several other species. Fig. 9, shows one case in Cheil. frondescens, 



I). See a'so PI. I. fig. 3. Cor, verae Japon. 



-!). Xereis Australis ; Amp. Bowcrbankii, Harv. p. L<7. t. XXXVII. 



