1898.] SPECIES OF MILLEPOEA. 249 



zoophytes that its presence is uot discovered until a fracture is 

 made. To give only one example to illustrate this point : — A 

 specimen in the Manchester Museum was named Millepora 

 intricata, and, on comparing it with the description of the species, 

 I thought at first that the name was correct. On breaking it into 

 two pieces, however, I found that the form it had assumed was 

 due to the fact that it had grown over a small piece of wood. 



In a still greater number of cases, however, the Millepores grow 

 upon the dead coralla of other Millepores or Madrepores or other 

 white corals, and then the difficulty of determining whether the 

 form of the specimen is due primarily to the liviug coral or to the 

 one on which it has grown becomes extreme. There is a large 

 specimen in the collection brought home from New Britain by 

 Dr. AVilley, of very irregular form, one part of which has a form 

 like that attributed to the species JM. plicata, another part to the 

 species M. verrucosa, but a broken knob shows quite clearly that a 

 part of this great mass has grown over a dead coral. It would 

 consequently be quite impossible to determine with any degree of 

 satisfaction to which of the already-described species it belongs, 

 unless every knob and projection were broken ofi to see whether 

 the dead coral extends as a basis through the whole piece. 



In the second place, the immense amount of variation in form 

 which occurs in large specimens of Millepora, and, indeed, in many 

 small specimens too, leads to verj^ great difficulties in the deter- 

 mination of species which have been described on form as the 

 principal character. In Dr. Willey's collection there is a series of 

 varieties of growth leading from a massive lamellate form to a 

 complicated branching and anastomosing form. 



A careful study of these skeletons, then, points very definitely to 

 the conclusion that the general form of the corallum of Milhpora 

 should be used, not as a primary, but as a very subsidiary character 

 in the description of species. 



The form assumed by the corallum must depend upon many 

 circumstances connected with the exact spot on which it grows. 

 If a Millepora embryo happens to become fixed on a large piece of 

 dead coral, it will form a large incrusting base, and such a base nearly 

 always gives rise to a lamellate form of growth ; if, on the other 

 hand, the embryo settles on a small stone or other object, lamellate 

 gro\vth is impossible, and the corallum will be ramified. 



The growth of the corallum must. also be influenced by the 

 propinquity of other corals. Its form must be adapted to the 

 space left between its neighbours on the crowded reef. Again, its 

 form must be modified by the depth of the water in which the 

 embryo happens to develop. As Duchassaing and Michelotti 

 pointed out long ago, Millejoora often grows in very shallow water 

 and is consequently unable to develop in height. Specimens that 

 happen to fix themselves on foreign bodies on the edge of the 

 reef at a depth of 5 or 6 fathoms can and do grow to a very great 

 length without impediment. 



It is also extremely probable that the available food-supply, the 



