2s6 DISCOVERY REPORTS 



The Other discrepancy Hes in Marshall's introductory words, where he summed up the specific 

 characteristics as 'distinctly bilateral'. However, his drawings (1883, PL 25, figs. 29, 30) clearly show 

 that the bilateral arrangement of autozooids in the cluster is also quite rudimentary in his rather small 

 specimen, and that, to cite his own words, ' it is difficult to make out any definite plan of arrangement 

 of the polyps '. 



Kolliker's specimen from the 'Challenger' and the statements of Kiikenthal and Broch (191 1) all 

 indicate a similar rudimentary bilateral arrangement in young specimens of U. huxleyi, and Marshall 

 evidently had a young stage of this species. 



The large specimen of 'U. gracilis' described from the 'Michael Sars' Expedition (Broch 1913) 

 with a total length of 547 mm. had no less than 25 autozooids crowded together in the cluster which 

 was almost globose (a pompon) because the polypiferous distal part of the rachis was extremely short. 

 No bilateral arrangement of the autozooids could be traced, nor was it possible to point out any 

 primary polyp because the axis did not run into the wall of any autozooid, but terminated in the 

 swollen rachis. 



In 1955 Dr A. R. Longhurst collected no less than 17 specimens of huxleyi in one haul with an 

 otter trawl off Bissagos I. (Port Guiana) on a muddy bottom in 600 m. depth. This collection was 

 generously placed at my disposal, and Dr Longhurst informed me that the comparatively large hold- 

 fasts ('stalks') of the specimens were rather strongly inflated and pulsating, when they came on deck. 

 This is the first time apparently that pulsation has been observed in an Umbelliila, and confirms the 

 view that the lower part of the stem is, and also ought to be termed, a holdfast for digging into the mud. 



Most of Longhurst's specimens measure about 27-28 cm. in length, three or four are a little shorter, 

 as small as 18 cm. The specimens with two exceptions were stowed together and crushed into a 

 container, so that it is impossible to count the autozooids in the clusters. Nevertheless, it can be seen 

 that these agree with the polyp clusters described by Marshall (1883) and Broch (191 3). Two 

 specimens (PI. i, fig. la, b) had been picked out and preserved in a special glass container; they give 

 an excellent impression of the specific characteristics. 



The smaller specimen {b) has 14 fully developed autozooids and a 15th bud in the central part of 

 the cluster. Eight of the autozooids constitute an outer (basal), somewhat irregular whorl, the others 

 are crowded within this whorl and rather irregularly arranged. No primary polyp can be traced with 

 any certainty, and with exception of the bud, the autozooids are of all but equal size. In the larger 

 specimen {a), 10 autozooids constitute the outer (basal), very irregular whorl, and eight other auto- 

 zooids of the same size are irregularly crowded within it showing no bilateral arrangement. It is 

 impossible to trace a primary polyp. The largest autozooids of the outer whorl (excluding the tentacles) 

 measure 17 mm. in length with a breadth of 3-5 mm. The autozooids of this specimen contain nu- 

 merous eggs, whereas no eggs were observed in the smaller specimen. 



Although of subordinate importance, it should be noted here that the tentacles of the preserved 

 specimens are comparatively short in most adult autozooids, measuring }, to h of the body length. In 

 a few polyps the tentacles are more extended, being as long as or even longer than the body. The 

 largest autozooids were found in the basal whorl. The dimensions are the same whether the specimens 

 were filled with eggs or were sterile adult autozooids. 



Symmetrical features are displayed with varying distinctness. In a few cases there is a strip without 

 autozooids along the dorsal side of the swollen rachis, where it is possible to discern the almost hair- 

 thin terminal part of the axis ending in the basi-dorsal wall of one of the autozooids, which is probably 

 the primary polyp, although not differing in any way from the other autozooids. In every richer 

 cluster, however, the upper part of the axis has been completely buried in the tissues of the rachis, 

 and the polyp-free dorsal strip has disappeared completely ; every trace of bilateral arrangement has 



