Simpson — A Revision of the Gorgoncllidae. 63 



Some of these — e.g., jia^idlum and elatu/ata — have, at times, appeared 

 under other generic names; but these will be discussed later under the 

 species in question. 



{h) Classification of the Species. 



In formulating a scheme of classification for these different species, and 

 also th(i large numlier of undescribed specimens which I have before me 

 for identification, two courses were available, either (1) to describe every 

 individual specimen, and name it on account of certain differences which may 

 or may not be inherent, or (2) to study the group as a whole, tabulate all the 

 points of difference in the various specimens, eliminate all variations, such as 

 occur in the same colony, reject all environmental modifications, and arrange 

 the specimens around some central type. The latter plan has been adopted 

 in the present work ; and for this reason it has been necessary to aliolish 

 several of the previously described species, not on account of their absolute 

 identity with formerly described species, but on account of the differences 

 wiiich obtain in these different forms having proved to be not greater than 

 differences appearing in an individual specimen. A very good example of 

 this is seen in the case of Scirpearia furcata. Such a procedure has been 

 possible in the case under consideration only on account of the large number 

 of specimens which it has been my privilege to examine ; and it is more than 

 probable that when a larger mass of material is available, it may still be 

 possible to diminish the number of species in this report. 



The characters on which the present classification are based are the 

 following : — 



(1) the number of main longitudinal canals, 



(2) the nature of the spiculation, and 



(3) the nature of the branching. 



These, of course, are not all of equal value; imt a reri/ rigifi .ir^tnmtion 

 may be made into two classes based on tiio number of nuiin longituilinal 

 canals. It has been found that in this group specimens have either (ur> or 

 four main canals. 



The nature of the branching wIumi it comes to be a question of "simple 

 or branchcii," as we have already pointed out, is of little value except in 

 certain well-defined species. This is very evident in such colonies as those 

 described under Scirpearia furcata. On the other hand, the very character- 

 istic mode of branching seen in Sciri)C'iria pectiiwta woulil seem to justify 

 its inclusion as a specific character. 



Scirpearia andamanensis and Scirpearia ramosa are also worthy of 

 consideration in this respect. 



