5i 



part of the zooecium is said to have a similar character; but specimens sent to the British 

 Museum (13.7. 10.4.) from Cullercoats, Northumberland, by Miss R. E. Roper 1 ), and apparently 

 referable to A. jiörosum, have an orifice which seems to be of the typical Arachnidium form ; 

 the appearance of an elongated tube being obvious only in those individuals in which there is 

 some eversion of the kamptoderm. 



The 'Siboga' specimens were growing on the surface of what seemed to be a fragment 

 of the axis of an Alcyonarian. I did not succeed in removing them sufficiently completely to 

 be able to describe the arrangement of the zooecia in any detail ; but there appears to be no 

 essential difference, in this respect, between the present species and Arachnidium irregulare. 

 In the existence of marginal crenulations, the present form closely resembles AracJinoidea ray- 

 lankesteri. The very large collar (figs 9, 10) distinguishes it from that species, and is the feature 

 referred to in the specific name suggested. 



When fully protruded the collar forms an elongated cone, the sides of which are but 

 slightly divergent. It is supported by strong ribs, each of which ends distally in a strong tooth 

 (tig. 10), which is not quite terminal but projects into the cavity of the collar, the distal edge 

 of which is everted in such a way as to form a marked angle with the rest of the structure. 

 In zooecia in which the peristome is not so long (fig. 8) the proximal end of the retracted 

 collar may be seen projecting into the general zooecial cavity. 



The specimens are not well enough preserved to allow me to describe the anatomy: 

 but I have evidence that the tentacles are numerous and not less than 1 6 in number. 



The measurements (in u) of the specimens figured are as follows : - - total length of the 

 zooecium to the left of fig. 8, including its narrovv proximal part, 2700 ; without this part, 750 ; 

 breadth of the zooecium (fig. 11), 650 ; length of the peristome, when completely retracted, 

 400 — 700 ; diameter of the same part, 120; length of the collar, 370 ; diameter of the collar 

 at its base, j^\ at its distal end, 190; of the proximal part of the zooecia, 40 — 50. 



A. protecta seems to be closely allied to A. ray-lankesteri. The marginal crenulations 

 of the zooecium are not quite so marked, or so regular as in that species, in which, moreover, 

 the peristome is less strongly chitinised and is not so quadrangular as in the 'Siboga' material. 

 Rousselet, in the paper cited above, figures no collar; and I have been unable to satisfy 

 myself of its existence in the specimens from Lake Tanganyika (Brit. Mus., 08. 5. 24. 9a, 9 b, 

 1 1 a — 1 1 c) described by that author. The differences in these parts seem to be a sufficiënt 

 justification for describing the 'Siboga' specimens as the representatives of a new species. 



Annandale 2 ) has recently described a form to which he gave the name of Platypolyzoon 

 investigatoris n. g., n. sp., from the telson of Squilla investigatoris, obtained off the S. W. coast 

 of Arabia, in 110 fathoms; and considers that it is allied to Arachnidium or perhaps to Arach- 

 noidea. In the great length of its peristome it resembles the latter genus, from which it differs 

 in the characters of the attached part of the zooecium. 



1) Roper, R. E., 1913, Dove Marine Laboratory, Cullercoats, Report for year ending June 1913, p. 53, PI. II, figs 2, la. 



2) Annandale, X., 191 2, Ree. Ind. Mus. VII, Part if} N° 13, p. 124. 



51 



