,4 DISCOVERY REPORTS 



I have already pointed out (p. 13) that the budding structures in Bougainvillia platygaster cannot be 

 regarded as blastostyles. Consequently, we have here a method of asexual propagation different from 

 the one met with in the two species just described. Nor can the polyps produced by B. platygaster 

 be compared with the actinulae of Hybocodon or Margelopsis for example. Actinulae are developed 

 from fertilized eggs, which have remained attached to the stomach wall of the medusa, whereas I 

 would emphasize again that in all the budding specimens of B. platygaster the gonads were immature. 



Although in B. platygaster medusa buds arise in three apparently different ways, it must be remem- 

 bered that an outgrowth on any polypoid structure of a hydrozoan is initially unmodified. Its future is 

 usually predestined by its position on the colony, which indicates whether it will develop into a 

 stolon, a hydranth, a blastostyle or a gonophore. Heteromorphic development, however, often occurs, 

 one of the most frequent results of which leads to the development of a stolon from a bud destined to 

 become a hydranth, and Berrill (1949) has demonstrated that heteromorphism can be due to 

 environmental factors. 



In B. platygaster, clearly, we have three examples of heteromorphism occurring within a single 

 species. The branched structures described as issuing from the corners of the stomach of the medusa, 

 each carrying a hydranth (sometimes two), are comparable with small hydroid colonies. 



In the specimen from St. 1581 (described under (2)) in which no hydranths are developed, the 

 whole ' colony ' issuing from the corner of the stomach has been transformed into a creeping stolon 

 adnate to the subumbrella of the medusa. The formation of a terminal hydranth is inhibited and 

 medusa buds issue directly from the stolon. These two structures invariably arise from the corners of 

 the stomach. 



Where outgrowths from the lateral walls of the stomach (described under (3)) have no possibility 

 of attachment to a substratum, no stolons and no hydranths are formed but only a cluster of medusa 

 buds on branched pedicels. These may be regarded as completely reduced polypoid colonies. 



Although the formation of hydranths by asexual budding from a medusa is contradictory to our 

 usual concept of the normal developmental cycle in hydromedusae, the budding in Bougainvillia 

 platygaster forces us to abandon the idea that the asexual offspring of a medusa must always be other 

 medusae. 



Family Pandeidae 



Merga rubra sp.n. 



(PI. II, fig. 4) 



Occurrence: St. 661. 2. iv. 31. 57° 36' S, 29° 54' 30" W. North of the South Orkney Islands. Net: TYFV 

 500-250 m. I specimen, the holotype. 



In spite of the poor condition of the single specimen, it must be described as a new species. The 

 umbrella is 7mm. high, including the slender and pointed apical projection; the diameter is about 

 4-5 mm. The walls of the umbrella are fairly thick. The stomach is about three-fifths as long as the 

 height of the bell-cavity and very broad ; the perradial edges are attached to the radial canals along 

 more than half the length of the stomach. The oral lips have disappeared. A broad conical apical 

 chamber leads upwards into the gelatinous apical projection. The gonads are mutilated ; apparently they 

 are smooth, but their size cannot be gauged ; they contain many fairly large eggs. The four radial canals 

 and the ring-canal are fairly narrow. There are two opposite marginal tentacles with very large conical 

 basal bulbs; ocelli not seen. Moreover, there are six rudimentary tentacles, or tenaculae, slender and 

 tenon-like, solid; two of them are opposite to two of the radial canals, the others are interradial, one 

 in each quadrant. Velum,? Colour: the stomach has a deep reddish brown colour, similar to that 

 of many deep-sea medusae. 



