792 C. FORSTER-COOPER. 



down to the distribution of the nematocysts on the tentacles, but unfortunately the preser- 

 vation of many of the specimens, especially those with smaller pol;y'ps, was too bad for minute 

 examination. The damage was probably done right at the beginning, the polyps owing to 

 their extreme delicacy being quickly spoilt in the short time between their capture by the 

 trawl and their preservation. 



We have left then a certain number of characters on which to found our species, all 

 of which are of a type in which we may expect, and usually find, great variation. They 

 may be briefly summarised as follows : 



1. Method of growth. 



2. Colour. 



3. Size of polyps, and their distribution on the corallum. 



4. Shape of polyp and its tentacles. 



.5. Size, shape, and distribution of spines, and presence, or absence, of secondary spines. 



Taken separately these characters are not of much use, for instance the growth of a 

 colony may be easily altered by outside influences, such as currents, temperature or locality. 

 Schultze quotes the case of Savaglia (Gerardiu), which in its normal condition grows as a 

 slender, well branched bush, a metre or more in height. In regions, however, where coral 

 fishing is carried on, the fishing nets continually dragged over the bottom break down these 

 tall colonies, and as a result forms of stunted gi-owth with thick bases, sending out a large 

 number of short branches, are found. If such a cause as this can alter the form of growth 

 of a colony it is extremely probable that the many natural physical differences of environ- 

 ment, which obtain in different localities, must act in the same way and produce different 

 forms of gi-owth. The same author also points out that from this very reason forms, which 

 have the same type of growth, must not of necessity be closely allied, since it may be the 

 result of convergence due to these causes. 



Of the value of colour we are in this group more or less ignorant, the onlj' specimens 

 obtained with a distinctive colouration being Aiitipathes rubra and Cirripathes anguina. 



The shape of the polyp in the groups under consideration shewed too great similarity 

 to enable any conclusion to be arrived at ; all were either round or oval or both on the 

 same corallum, but the younger polyps were frequently found to be more oval than the more 

 fully grown ones. The height of the polyp is also so much a matter of relative contraction 

 that too much stress must not be laid upon it. There is, however, a certain difference in 

 absolute size between different forms, by which several groups may be roughly arranged. 



Thei'e is left, then, the spines — their shape and arrangement — and here there seems to 

 be a character of more promise. However, they will only serve to differentiate between 

 species, there being no type of spine entirely confined to one group so as to mark off that 

 group from all others. Also two forms differing in method of growth, size of polyp and 

 other particulars, may yet have very similar spines, so that it is only by combining all 

 these characteristics that we can form a working classification. Whether it will ultimately 

 prove a natural one is doubtful. 



All the figures are magnified 8 diameters unless otherwise specified. I have to thank 

 Mr Crossland for three specimens from E. Africa. 



