I-':.. in tin-- i. tl-li' it will lie -n that all the species of the Ceratoisis group, except 

 ('. nwl>i and (\ .//(///(/(//.-/I/ from Fiji and C. rarmtsa from the Cape, are found 

 mirth <>f the K-jimtor, whereas all the species of the Primnoisis group arc fouud in 

 tin* Antarctic seas or south of the Equator. Ceratoisis ninumi is in many respects 

 an intermediate form, and is perhaps more closely related to the Primnoinis than 

 the Ceratoisis group ; and Ceratoisis spicata is also an intermediate form with closer 

 .I trinities to the Ceratoisis group. 



FAMILY PRIMNOIDA 



This family is represented by two genera, Thouarella and Primnoella, both 

 belonging to the sub-family Primnoidinte. 



TUOUARKLLA ANTARCTICA (12). 



Primnoa tmtarttiea, Milne-Edvarda, Hut. nat. Corall., L (1857), p. 140 [Vulenciennea Atlas, Toy. 



Vcnna (1846), Zooph., pi. 2, fig. 2]. 

 ThouartUa aittaretica, Wright and Studer, Chall. Rep. Alcyon. (1889), p. 65 ibiqut citato. 



(Plate II., figs. 19, 24.) 



Localities: 1. W.Q., February 28, 1903. No. 6 Hole. 130 fathoms. 



2. Off Coulman Island, January 13, 1902. Dredge. 100 fathoms. 



3. McMurdo Bay, February 8, 1902. 96-120 fathoms. 



4. Station 270, March 4, 1904. (Just within Antarctic Circle.) 



254 fathoms. 



The specimens all belong apparently to one species. Only one specimen with 

 the base broken off was obtained in the first locality (February 28, 1903), and this 

 I have no hesitation in placing in Valenciennes' species. The specimens from 

 off Coulman Island, locality 2, were treated with picric acid before preservation 

 in alcohol, and this has changed the character of the spicules and thereby the 

 size of the calices to such an extent that the determination of their species is 

 rendered uncertain. The specimen from McMurdo Bay is only a fragment, but 

 it is well preserved, and the calices are identical in general characters with those 

 of the specimen from locality 1. 



All the specimens have the characteristic bottle-brush (" goupillon ") mode of 

 branching. Several of them have the main axis quite simple, the twigs, as in the 

 type specimens, being much more slender than the axis, but the specimen from 

 locality 1 shows two dichotomous branchings at the base of the axis where the 

 colony is dead and bare, and one specimen from Coulman Island has the axis 

 bifurcated so that two " goupillons " spring from the same axis. 



The heights of the specimens from Coulman Island are 450 mm. (the specimen 

 with bifurcated main axis), 175 mm. (a small specimen, but the only one with base 

 of attachment), 200 mm. and 150 mm. The height of the specimen from the 



.in 2 B 



