THE SPECIES OP THE GENUS MILLEPORA. 135 



tially a shallow-water genus, living in abundance in most of the coral- 

 reefs, and not occurring in greater depths than 15 fathoms. 



The form varies immensely. It may be broadly lamellate or densely 

 branched, or anastomosing, or it may form thin incrusting plates on 

 dead corals. In all large collections of Millepores series of intermediate 

 forms may be found between all the most prominent types. 



The difficulty of defining and describing the species of this genus has 

 been commented upon by several authors. Dana, for example, says 

 " There is much difficulty in characterizing the Millepores on account 

 of the variations of form a species undergoes and the absence of any 

 good distinctions in the cells. The branched species are often lamellate 

 at the base, owing to the coalescence of the branches, and the lamellate 

 species as well as the branched sometimes occur as simple incrustations." 

 My own investigations confirm and amplify Dana's statement on this 

 point. 



Notwithstanding these difficulties a large number of species of the 

 genus have been described. In the writings of the older naturalists 

 many species were described which have since been relegated to other 

 classes of the animal kingdom, and in paleeontological literature we find 

 many species of fossil corals referred to the genus on erroneous or very 

 unsatisfactory grounds. 



Apart from all these, which may be left out of consideration in this 

 paper, no less than 39 species of the genus Millepora have been 

 described. 



The characters which have been used for determining thes3 species 

 are: — (1) The form of the corallum. (2) The size of the pores. (3) The 

 degree of isolation of the cycles. (4) The presence or absence of 

 ampulla?. (5) The texture of the surface of the corallum. 



(1) The Form of the Corallum. — This feature is even more unsatis- 

 factory than I anticipated at the beginning of my investigation. In 

 the first place, attention has been called by Dana, Duchassaing and 

 Michelotti, and others to the fact that Millepora grows in an incrusting 

 manner on many objects, and thereby assumes the form of the object 

 on which it grows. It is quite easy to distinguish such forms as 

 incrusting forms when they have only partially covered such objects as 

 the horny axis of a Gorgonia, a glass bottle, or an anchor • but in many 

 cases the object is so completely overgrown by Millepore and other 

 marine zoophytes that its presence is not discovered until a fracture is 



