278 Mr Bernard, On the Unit of Classification 



stony corals. I have now worked long enough with it to be 

 convinced not only that it is a workable symbol but that it is one 

 which promises to be an instrument of work of real value. But 

 my experience has taught me that it is almost useless at this 

 stage to discuss, and quite useless to wrangle over, mere sug- 

 gestions. The first thing is to establish the need for the change ; 

 it will then not be long before we have several earnest attempts 

 made to discover a satisfactory formula. Time will select the 

 best, the best being that which, in practice, satisfies most com- 

 pletely the new demands. 



Supplement. 



{A brief description of the method of designating the varying 

 forms suggested by the author; for some important details he is 

 indebted to Professor Jeffrey Bell, of the British Museum, and to 

 Professor Weldon of Oxford.) 



A few examples will serve best to explain the system. 



Goniopora New- Guinea \. 



[Port of Doreh, north-west corner of Great Geelvink Bay, 



Dutch New-Guinea; voy. de L' Astrolabe; Paris Mus.] 

 Syn. Goniopora pedunculata Q. & G. etc., etc. 



Goniopora Tonga Islands f. 



[Tongatabu. Tonga Islands ; coll. J. J. Lister ; Brit. Mus.] 



Goniopora Vicenza -^ T . 



[S. Urbano, near Montecchio-Maggiore, Vicenza ; Eocene 



"Calcaria Grossolana".] 

 Syn. Porites ramosa, Catullo [non Reuss] Corals of the Venetian 



Alps etc., etc. 



The first line in each case is the designation, for the purposes 

 of naming and reference. It consists of (1) the genus name, 

 (2) the district, in which the locality occurs, (3) numbers 

 arranged like a fraction. 



The genus name is quite suitable for the corals, at least in 

 my own case, because as I am practically monographing the whole 

 group I can satisfy myself that the genus is fairly stable. I find 

 it is the recently established genera which cannot be relied upon, 

 and which have to be suppressed as our collections increase. An 

 old genus, if still accepted, and from which other new genera 

 have been separated, is generally safe. As a rule, however, if 

 there were great doubts, we might adopt some higher divisional 

 name, for instance the family, with the genus suggested with a 



