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



of Physonects, which I coruinunicated to the Medicin. Naturwiss. Gesellschaft in Jena 

 on July 8 and November 25, 1887 (95, pp. 38-42). I then gave the definition of 

 twenty-five different genera (ten of which were new), and disposed them in eight 

 families, four of which were new. Four families had already been distinguished by 

 Huxley (9, p. 70), viz.: — (1) Apolemidse; (2) Stephanoinidae ( = Agalnnda3, Brandt); (3) 

 Physophoridae (sensu restricto) ; and (4) Athorybidse ( = Anthophysidae, Brandt). The 

 four new families of my system are — (l) Circalidaa ; (2) Athoridaa ; (3) Forskalidae ; and 

 (4) Nectalidae. 



All authors up to this time have employed for this order the name Pkysophoridse, 

 given by Eschscholtz (1, p. 139). But this name has become unserviceable, since it is 

 used with no less than four different meanings (95, pp. 38, 41). Huxley applies it to 

 all Siphonophoraa, with the single exception of the Calycophoridae (9, p. 67). The 

 majority of authors employ the name in the sense of Eschscholtz, comprising not only 

 our Physonectae, but also the Cystonectae (fifth order). Chun first separated the latter 

 from the former, and proposed the term Physophoridae for the first, and Pneumato- 

 phoridae for the second (86, p. 1168). But, besides, the special family which I call 

 Discolabicke (XVI.) is generally named Physophoridae. To avoid this general confusion, 

 I propose for this order the name Physonectae. 



Nectosome. — The swimming apparatus is composed in all Physonectae of an apical 

 float or pneumatophore placed at the top of the trunk, and of a variable number of 

 nectophores or swimming-bells, arranged in a variable manner around the trunk ; rarely 

 these nectophores are replaced by paddling bracts (Athoridae, Anthophysidae). The 

 Physonectae have therefore combined in their locomotive apparatus the active swimming- 

 bells of the Calyconectae and the passive hydrostatic float of the Cystonectae and 

 Disconectae. They agree in this combination with the Auronectae, but differ from these 

 in the absence of the aurophore as well as in the simple tubular trunk and the structure 

 of the siphosome. The physiological function of the two portions of the nectosome is very 

 different ; the hydrostatic float is comparable to the swimming-bladder of the fishes, and 

 enables the Physonectae to rise or descend ; whereas the active movements of the necto- 

 phores, or the paddling bracts which these replace, effect the voluntary locomotion in 

 every direction. 



Pneumatophore. — The float filled with air, rising from the top of the axial trunk, 

 is in all Physonectae relatively small ; much smaller than that of the Auronectae and 

 Cystonectae ; it rarely attains the size of a nectophore (Circalidoj) ; usually it is scarcely 

 half as large, or less. Whilst the physiological importance of the float rests in its 

 hydrostatic function (as " swimming-bladder "), its morphological nature is explained by 

 its development; the pneumatophore is the modified umbrella of the original Medusa, 

 the manubrium of which is the axial trunk (or the first siphon). (Compare above, p. 11-) 

 Its form is usually ovate, spindle-shaped or pyriform, sometimes subspherical, at other 



