338 THE VOYAGE OF H.M.S. CHALLENGER. 



terminal branches, an odd median ampulla and a pair of lateral horns (fig. 1, a). If 

 this explanation be right, Angela would bear the same relation to Epibulia as Canno- 

 physa does to Aurophysa. 



Family XXIV. Physalid^e, Brandt, 1835. 

 Physalidm, Brandt, Prodromus descriptionis, &c, SJ5, p. 36. 



Definition. — Cystonecta3 polygastricae with a short inflated trunk of the vesicular 

 siphosome, which is horizontally expanded along the ventral side of the subhorizontal 

 gigantic pneumatophore. Cormidia in a multiple series along the ventral side of the 

 trunk, usually loose. Pneumatosaccus very large, sometimes with a chambered dorsal 

 crest, without radial septa and hypocystic villi. 



The family Physalidse comprises the largest and most interesting Cystonectse, which 

 surpass all the other Siphonophorae in the enormous size of the inflated pneumatophore 

 and the vigorous tentacles, the splendid colours of the peculiar siphosome, and the 

 dangerous poisons of the urticating organs. Since some forms of this interesting 

 family, and especially the largest of all, the Atlantic Caravella maxima, occur in great 

 swarms floating on the surface of the ocean, they have been well known to travellers and 

 sailors for a long time. In maDy voyages they are mentioned under the popular names 

 " Caravella, Galera, Fregatta, Ship of Guinea, Portuguese Man-of-War," &c. Never- 

 theless the natural history of the Physalidse has hitherto been very imperfectly studied, 

 and there remains much to be done towards a complete knowledge of them. 



The genus Physalia, generally accepted as the only one of this family, was established 



in 1816 by Lamarck (90) for those Caravellse which 0. F. Midler and Gmelin in the 



preceding century had described as Medusa caravella, Linnd, or Holothuria physalis. 



Neither the name Salacia, employed for them by Linne in a few editions of his 



Systema naturae (1756, 1, p. 158), nor the name Arethusa, proposed by Patrick 



Browne in 1789, have yet any application (compare above, p. 330, and below, p. 349). 



While Lamarck distinguished five different species, with very insufficient definitions, 



Eschscholtz accepted only three, the common Atlantic Physalia caravella (the largest of 



all, with numerous large tentacles), the smaller Physalia pelagica (from the Southern 



Atlantic and Indian Oceans), and the larger Physalia utriculus from the Pacific ; the two 



latter species have only a single large tentacle, and differ mainly in the form of the 



pneumatophore. Recently Chun, in the latest paper on Physalia (83, p. 557), unites 



these two latter forms, accepting two species only of this genus — Physalia caravella, with 



many large main-tentacles, from the Atlantic, and Physalia utriculus (including Physalia 



pelagica) from the Indo-Pacific Ocean. All recent authors have accepted the genus 



Physalia as the only type of this family. 



