The earlier writers usually included the species now allotted to the family Melitodidce 

 in the old genus Isis, thus involving the classification of the I sides in considerable confusion. 



KOLLIKER (1865) with his usual clear insight, separated the " Melithceacea" and Isidinae 

 as distinct subfamilies of his Gorgonidae, using the characters of the axis cylinder. 



GRAY (1870) established the families Isidce, Mopseadce, Acanelladce and Keratoisidce, all 

 of which are now included in the single family Isidce. 



STUDER (1887) gave a careful diagnosis of the family Isidce, and divided it into the 

 subfamilies Ceratoisidints, Primnoisidince and Isidince. Later (1889), in the Challenger Report, 

 WRIGHT and STUDER defined the family as follows : 



"Colony consisting of a simple or branched axis. The axis consists of calcareous and 

 horny (internodal and nodal) regions; the branches when present arising from either the nodal 

 or internodal regions, sometimes anastomosing; the axis solid or hollow, smooth, fluted or 

 echinulate. The base of the axis calcareous and attached". 



To avoid any confusion of certain species of the Melitodidce with the Isidce, this definition, 

 although substantially correct, may be modified for the purposes of the present work as follows : 



Gorgonacea with an axis composed of alternating calcareous and horny segments, the 

 former being amorphous and not composed of a mass of agglutinated spicules, and the latter 

 entirely horny without the admixture of definite calcareous spicules. 



Several authors have proposed breaking up this family into subfamilies; but with the 

 increase of our knowledge of the group this becomes increasingly difficult. GRAY (1870) went 

 to the extreme of placing the species now included in the Isidce into four distinct families, 

 viz. Mopseidce, Acanelladce, Keratoisidcz and Isidce. 



VERRILL (1883) separates all but the genus Isis into a family Ceratoisidce, which therefore 

 includes GRAY'S Keratoisidee, Acanelladce and Mopseidce. STUDER (1887) divides the family into 

 the subfamilies Ceratoisidints^ Primnoisidince (including Mopseidce) and Isidince. In their Report 

 on the Alcyonaria of the Challenger Expedition (1889) WRIGHT and STUDER substitute the name 

 Mopseince for STUDER'S subfamily Primnoisidinee, and base the subfamilies of the Isidce on the 

 characters of the spicules. 



The trouble with this arrangement is the intergrading features between the Mopseince 

 and Ceratoisidince . In his discussion of the genera Ceratoisis and Primnoisis, which belong to 

 the Ceratoisidince and Mopseince respectively, HICKSON (1907) holds that these genera are not 

 distinct, and proposes including both in the genus Ceratoisis. He says: 



"The study of many specimens belonging to this family has convinced me that this 

 subdivision is unnecessary and inconvenient. The many variations of spicule characters that are 

 found in the species of a single genus render these structures unsatisfactory for the purpose of 

 wide systematic differentiation. If we take a single species from each of two of the subfamilies 

 and compare them, the differences observed in the character and arrangement of the spicules 

 may seem to be of a higher rank than the usual differences between genera; while, on the 

 other hand the examination of a large number of species of the same two genera will reveal 

 so many of the intermediate conditions as to render the separation of the genera, on spicule 

 characters alone, impossible. This kind of difficulty is particularly well seen in the case of the 



