Vlll 



included four fine species of Pennatula and one accidentally included species 

 of Ptero'eides (from only 56 fathoms) ; the Littoral Collection has one accidentally 

 introduced representative of Permatida (from 609 fathoms), and sixteen species 

 of Ptero'eides. 



{b) A glance at the collection as laid out on the shelves shows at once as 

 the most conspicuous feature the large number of species (61) of Dendronephthya 

 (or Sponr/odes) and the large number of these that we have had to label as new. 



(c) To a lesser extent the same may be said in regard to Siphonof/07r/m 

 and Ptero'eides. 



(d) The fine representation of Muriceids is noteworthy. Thus we have 

 described six new species of Echinomuricea, three of Echinogorgia, two of Acis, 

 two of Miiricella, and two of Eumuricea. 



{e) Perhaps the most characteristic feature is the representation of what 

 we feel inclined to regard as annectent types related to Alcyonids, Nepthyids, 

 and Siphonogorgids. We refer to the genera Stmlerlote.'^, DactDlonepMhya and 

 Cactogm'gia. 



Some Taxonomic, Notes. 



It seems to us that the species of Dendronephthya or Spongodes are in a 

 state of flux. Some are sharply marked off from their neighbours, but most 

 are separated by an ensemble of somewhat minute features. It may be that 

 further investigation will enable us to group the species around centres, but 

 the time for that has not yet come. It would be interesting to study some of 

 the species statistically, but that would require to be done on the spot. 



We have studied a large number of specimens which are referable to 

 previously described species of Siphonogorgia and Chironephthya ; we have 

 given a fair trial to all the characters which have been advanced as distinctive 

 of these two genera in contrast to one another, and we have found that none 

 is satisfactory. We agree with Kiikenthal that C'hironej)hthya should be merged 

 in Siphonogorgia. 



In certain cases, such as Stereacanthia and Cactogorgia, it seems to us very 

 difficult to decide between the sub-families Spongodintie and Siphonogorginse, 

 but we have followed Bourne's characterisation of the Siphonogorginaj as 

 having " Canal walls densely filled with spicules ". It seems to us that the 

 contrast between the stem-canals in a type like Ennephthya rosea and in 

 Stereacanthia is a useful criterion, though annectent forms like Stereacanthia, 

 Cactogorgia, Agaricoides and Dactylonephthya help to bind Nephthyid and 

 Siphonogorgid types together, and have led us to prefer Bourne's division of 

 Nephthyid;e into two sub-families Spongodin^e and Si2:)honogorgin0e to Kiiken- 

 thal's recognition of two families Nephthyidae and Siphonogorgiidae. 



After a comparison of sevei'al specimens of Kero'eides we have come to the 



