434 DISCOVERY REPORTS 



C. areolatus (Fig. 54 A) and C. rectilinearis (Fig. 52 A) it is frontal, and in C. asym- 

 metricus asymmetrical. Runners have not been seen in typical C. bicornis, C. bicornis 

 var. compacta, or C. tricornis (Fig. 46 B). Where there is no rootlet the chamber is 

 usually inconspicuous or absent, but in C. retiformis (Fig. 47 H), C. latus, C. atlanticus 

 and C. rectilinearis a well-developed axillary chamber with runners is present whether 

 there is a rootlet or not, and in C. angustus (Plate XII, fig. 3) the axillary chambers give 

 rise to secondary branches (see p. 436), axillary rootlets being rarely formed. The forma- 

 tion of axillary secondary branches is occasionally seen in most forms possessing lateral 

 secondary branches, but is not usual except in C. angustus. 



The lateral rootlets also originate from chambers (e.g. Fig. 51 D). Except in C. 

 tricornis (see Fig. 46 C and p. 451) and some young colonies (p. 441) these are placed 

 basally and laterally in the angle between the narrow proximal end of one zooecium and 

 the wider distal end of the preceding one (Figs. 47 I, J, 51 D). The secondary branches 

 arise from these chambers, and a single chamber may produce both a rootlet and a 

 secondary branch. In C. giganteus, C. lewaldi, C. angustus, C. retiformis, C. latus, 

 C. atlanticus, C. rectilinearis and C. reticulatus the lateral chamber forms a runner 

 (Fig. 47 I, J). In C. tricornis the chamber is prolonged distally but tapers gradually, no 

 part of it being recognizable as a tubular runner (Fig. 46 C). 



These arrangements are not peculiar to Camptoplites, for in a specimen of Bugula 

 (Dendrobeania) birostrata Yanagi and Okada (1918, p. 420) from Misaki (21. 11. 7. 3, 

 500-600 fm.) the rootlets forming cross-connexions (cf. Silen, 1938, p. 237) arise from 

 chambers with runners, like those of Camptoplites. The chamber figured by Yanagi and 

 Okada has a mere protuberance in the position of the runner, but in the British Museum 

 specimen runners are well formed. 



The statement that the operculum is differentiated in Camptoplites is true, but per- 

 haps misleading. An arc marking its edge can be seen in many zooecia, but it is very 

 faint, far less distinct than that of Cornucopina for instance, and in many zooecia it 

 cannot be detected at all. 



Avicularia. The avicularia of Camptoplites are remarkable for their frequent pro- 

 fusion, diversity of form and length of stalk. The youngest stalks contain much tissue, 

 and do not collapse or contract in preservation. In older stalks this tissue gradually 

 disappears, and these may either contract longitudinally, when preserved, producing 

 fine transverse striations as seen by Busk (1884, p. 41) and Calvet (1909, p. 9), or they 

 may collapse and become crumpled, as described by Harmer (1926, p. 453). Part of a 

 stalk may collapse while another part of the same stalk contracts (see for example C. 

 bicornis var. quadriavicularis, slide 267 B 4 from St. 160). Contracted stalks give a 

 strong superficial impression of muscular action, but specially preserved material 

 would probably be necessary to settle, in such instances, whether contractile tissues 

 were present, as stated by Busk and Calvet. Harmer did not find muscles in the 

 avicularian stalks of C. lunatus, and throughout the genus the stalks of many apparently 

 healthy avicularia contain little or no tissue of any kind. It is thus clear that musculature 

 in the stalk cannot play an essential part in the functioning of these avicularia. All these 

 questions may have to wait until living material can be observed. 



