506 Mr. H. M. Bernard on the 



its younger stages. Outside it streams down over the stalk, 

 not only thickening it, but expanding its base of attachment. 

 If the slope of the outer surface of the cup does not permit it 

 to run down the stalk, it may either merely thicken the 

 corallum under the layer of intercommunication of the polyp- 

 cavities, or it may even hang down like aerial rootlets 

 seeking to attach themselves to the substratum independently 

 of the original stalk. 



Besides giving the corallum the necessary strength to stand 

 erect, this flow of matter is protective. It threatens every 

 parasite which endeavours to gain a foothold. The Balanids 

 have found it necessary to develop elaborate fringes of bayonets 

 to keep off the advancing tide. These are often successful on 

 the upper surfaces, where the flow of the coenenchyma is not 

 so strong ; but such defences are usually of no avail on an 

 under surface. A i>a/o7iMS attached to the undcrsurface is soon 

 engulfed by the coral-substance. In some cases, e.g. on per- 

 pendicular surfaces where the flow is also great, the Dalanus 

 manages somehow to keep itself from being submerged ; but 

 the efforts of the coral to get rid of the plague are evidenced by 

 the length of the finger-like processes, in the ends of which the 

 JJalani rest secure. They appear to have risen on the coenen- 

 chyma as it strove to surmount the edges of their bayonetted 

 plates *. 



Factors in the Growth of the Corallum. — We have, then, a 

 stalked corallum, with edges standing up to form a cup, or 

 standing out to form a disk, or hanging down to form a hemi- 

 spherical mass, and we have two factors to account for the 

 further growth of the corallum: — (1) The typical method of 

 budding; (2) the flow of the material building up the coenen- 

 chyma, this flowing being especially marked in Turhinaria 

 owing to the great abundance of this tissue. Before showing 

 how these primitive forms of the coralla become variously 

 modified by these two typical elements of increase, it is 

 necessary to describe a third, somewhat irregular, factor, viz. 

 an adventitious bud-formation. 



Adventitious Budding. — In nearly all Turbinarian coralla 

 with uneven surfaces the coenenchyma seems to accumulate in 

 the valleys as it does in the bottom of the early cups, sub- 



* A very beautiful con-elation exists between the size of the teeth on 

 the plates of the Balunus and the echinulations of the coenenchj'ma. The 

 line echinulations in the Turbinarians are met by tine teeth on the 

 Ualanus ; the long echinulations of the Astreeopora are encountered by 

 correspondingly long bayonets on the plates of the Balanus. 



While on the subject of parasitic or attached organisms, I may mention 

 that many infesting sponges " imitate " exactly the colour of the corallum, 

 and sometimes also the polyp-cavities in the size of their oscula. 



