l68 THE INVERTEBRATA 



centre a more or less regular vertical rod, the columella. The de- 

 pression always tends to become filled up by the secretory activity of 

 the general surface of the basal disc, but the building up of the theca 

 and septa keeps pace with this. It is difficult at first to realize that this 

 is an exoskeleton and that in a massive structure like a brain coral the 

 actual living tissue is a mere film on the surface of a great hemispheri- 

 cal mass of calcium carbonate which it has secreted. It is not sur- 

 prising to learn that such colonies with a diameter of a yard or more 

 have a life span of a hundred years or so. 



With regard to the actual mechanism of lime secretion the view 

 most generally held is that illustrated by Fig. 138 F, which shows a 

 coral larva which has fixed upon a piece of cork. The skeleton as 

 shown in a section is, when first laid down, a series of spheroidal 

 masses of calcium carbonate, which thus appear to be a secretion of 

 the ectoderm cells, issuing from the cells as a solution and immedi- 

 ately crystallizing out as irregular masses. Another suggestion is that 

 ammonium carbonate excreted by the coral meets the calcium salts 

 of the sea water and carbonate of lime is precipitated round the ecto- 

 derm ; and still another, that calcium carbonate is stored up in the 

 ectoderm cells and when the cells are full they drop out of the 

 epithelium and are added to the skeleton. 



Coral colonies exist in the most diverse shapes and forms (Fig. 139), 

 from the slender tree-like colonies of many Madrepora to the massive 

 rounded forms like Pontes. Each colony is formed from a single 

 planula which settles down and forms a polyp. From this first in- 

 dividual the hundreds of thousands of polyps in a large colony are 

 formed by division or gemmation. An example of division is given 

 in Fig. 138 D. In such a case when the polyp has reached a certain 

 size the oral disc becomes elongated in the direction of the long axis 

 of the mouth, tentacles and mesenteries increase in number, and 

 finally a transverse constriction divides first the mouth, then the disc 

 and lastly the whole polyp. The division of the polyp is followed by 

 that of the theca. In the Meandrine corals (brain corals) the polyp 

 elongates enormously and the mouth divides but not the theca, and 

 so we get the curious thecae running more or less parallel to each 

 other which recall the convolutions of the human brain. In Lopho- 

 helia (Fig. 138 E) division is equal, but while one of the polyps re- 

 sulting from it continues to grow the other marks time; the axis of 

 growth changes sides at each division and the result is a colony showing 

 cymose branching. 



In Fig. 138 B it is shown that part of the coral polyp overlaps the 

 theca. It is this extrathecal zone which gives rise to young polyps 

 when a colony is formed by gemmation (Fig. 138 C). The bud and the 

 parent remain connected by their extrathecal portions, and this con- 



