Investigator in the Indian Ocean, Part II, 1909, p. 265 et seq. merely name the family. 

 They distinctly recognize the great difficulty and perplexity attending the work of the systematist 

 in this family, saying (p. 267), in reference to their new species Nicella pustulosa: 



"It is a matter of no small difficulty to distinguish between Nicella, Goi'gonclla and 

 Verrucella. Distinctions based on spicules alone are very unsatisfactory in this group because 

 the spiculation varies at different levels and transition forms are so numerous and varied that it 

 is sometimes almost impossible to distinguish between double spheres, double stars and double 

 clubs, each in turn passing gradually to double spindles". 



Again, on page 269 : 



"The system of classification which at present obtains in regard to the Juncella group 

 of gorgonellids, including Juncella, Ellisella, Scirpearca and Scirpcarella is far from satisfactory. 

 In fact it is a debatable question whether these should be ranked as separate genera. Many of 

 the species which have from time to time been described have unquestionably been established 

 on young colonies, and, in addition to this, the characters which are taken as diagnostic, e. g. 

 arrangement and retractility of Verrucae, vary so much in individual specimens that little or no 

 importance can be attached to them". 



These writers content themselves with a table, giving a comparison in numerous details 

 of the several specimens collected by the Investigator, without attempting to name them. 



The present writer, although profoundly impressed with the e.xtreme difficulty of the 

 problem, does not feel justified in turning his back on these perplexities. The amount of 

 material belonging to the Gorgonellidie secured by the Siboga Expedition is so considerable, 

 representing some 2 1 species, that an attempt will be made to straighten out as many of the 

 difficulties as may be. While entire success is not to be hoped for, some progress toward a 

 reasonably correct generic classification should be possible. 



First, however, it is nece.ssary to more clearly define the main types of spicules that 

 present, after all, the best basis of classification in this, as in other groups. 



As already indicated, nearly all of the spicules which present characteristic features of 

 value for our purpo.se are "girdled spicules". The basic form from which nearly all of the 

 others are derived is the girdled spindle, showing at its middle a smooth impressed zone, free 

 from verruca;. This form differs from the typical spindle only in the fact that this zone is more 

 conspicuous in the girdled form. The typical spindle is terete in form and its surface is orna- 

 mented with more or less regular whorls of tubercles. Between the whorls are comparatively 

 bare zones, and such a zone is very constantly seen near the centre of the spicule. It is only 

 when such a zone is comparatively large and conspicuous that a "girdled spindle" is [produced. 

 This basic form is modified in the following ways : 

 i^'. Symmetrical forms 



a. The girdle divides two similar parts of the original spindle in such 

 a manner that each part bears Verrucae and is terete in outline. 



This produces the D o u b 1 e .S p i n d 1 e. 



b. The girdle divides two similar parts each with its outer end some- 

 what turgid and armed with Verrucae, resulting in the. . . . . Double Club. 



