THE GENUS ALVEOPOEA WITH THE FATOSITID^. 499 



In fig. 2 the original polyp has given rise to others. The 

 skeletal divisions between the component polyps of the young 

 colony are, to all appearance, wholly formed by the meeting and 

 fusion of septal ingrowths, while their outermost walls are of 

 epitheca. The rim of the epithecal cup is bulged and bent out- 

 wards round each polyp. 



In the third stage (fig. 3) the process of adding new polyps 

 has been continued further, the skeletal and purely septal walls 

 between the polyps rising high up in the centre of the cup, 

 considerably above the level of the rim of the epitheca. 



In figs. 2 and 3 we get further insight into the method of the 

 early formation of the skeleton of the bud. A low skeletal arch 

 rises off the epitheca, between which and the epitheca is the 

 minute fossa of the young bud, while on the outer side of the 

 arch the septal spines of the parent project. This arch must be 

 explained as due either to the meeting and fusion of two septal 

 infoldings, or of two other infoldings which are not septa. This 

 latter alternative seems to me quite unnecessary : the septal 

 infoldings are there, and the fusion and modification of two 

 adjoining septa to form the arch is easy to understand. We are 

 however, fortunately not left without some evidence of this 

 important morphological conclusion, viz. that the dividing walls in 

 4-lveopora are due to the interlacing of septal spines. 



1. A survey of a great number of specimens belonging to different 



species shows that a very intimate connection exists between 

 the septa and the walls ; if the septa are thick, the walls are 

 thick and solid: if the septa are wiry and thin, the walls 

 are delicate and fragile. 



2. The lattice-like character of the intervening walls is common 



to every species, and is therefore a primitive characteristic 

 of the genus. This is just what would arise in walls built 

 up of septal spines fused together in one plane. 



3. The septal spines in any calicle project as frequently from the 



transverse as from the vertical bars of the latticework, which 

 makes it improbable that these vertical bars are septa and 

 the transverse synapticulse. 



4. If a great number of the minute arches of young buds be ex- 



amined under a pocket-lens in a well-cleaned adult stock, 

 while it is true that their homology with septal spines is 

 most frequently as much obscured as in figs. 2 and 3, yet 



