REPORT ON THE SIPHONOPHORvE. 333 



in the North Pacific, have been incompletely described — Epibulia chamissonis lay 

 Eysenhardt (77, Tab. xxxv. fig. 3), and Epibulia erythrophysa by Brandt (25, p. 34). 

 Unfortunately the excellent figure of the latter species, which Mertens in 1817 had drawn 

 from life, has never been published. It is much to be lamented that this, as well as 

 all the other wonderful drawings of Siphonophorse, which Mertens had executed with the 

 most admirable accuracy, have never found their way into literature. Comparing Mertens' 

 figure of Epibulia erythrophysa with a similar splendid Cystonect which I myself 

 observed in Ceylon (1882), I have no doubt that both species belong to one and the 

 same genus. The accurate examination of this Indian Epibulia ritteriana (PI. XXII. 

 figs. 6-8), and their comparison with the closely allied Physalidse, has led me to 

 the opinion that this genus represents a new family, intermediate between the latter and 

 the Rhizophysidse. Another genus of this family may be perhaps Angela cytherea of 

 Lesson, which, however, is too imperfectly examined (3, p. 496, pi. ix. fig. 1). 



The Epibulidse agree with the next allied Physalidse in the bag-shaped form of the 

 short and wide trunk of the siphosome. But the large pneumatophore does not extend 

 along the dorsal side into the cavity of the vesicular trunk ; it occupies rather the apical 

 half of the corm, whilst the siphosome occupies its basal half. The Epibulidse agree in 

 this respect with the other Cystonectse, and differ essentially from the true Physalidse, 

 with which I had united them (in 1887) in my System (95, p. 46, Genus 73). A further 

 consequence of this important difference is that the main axis of the large pneumatophore 

 stands subvertically or obliquely inclined in the Epibulidse (as in the other Cystonectse), 

 whilst it is subhorizontal in the Physalidse ; the stigma, therefore (or the apical air-pore), 

 occupies in the former the superior pole, in the latter the anterior pole of the floating 

 corm. A further difference is, that the ventral line of the shortened trunk, from which 

 the budding cormidia arise, is circular or rather spiral in the Epibulidaa, straight in the 

 Physalidse, and that a complete corona of palpons surrounds the base of the float in the 

 former, but is wanting in the latter. 



The whole form of the corm, as well as the structure of its single parts, is in the 

 Epibulidse very like that in the Cystalidse (PI. XXII. figs. 1-5) ; but the important 

 difference between them is that the former are polygastric, the latter monogastric. It 

 is very probable, however, that the Epibulidse have arisen directly from the Cystalidse 

 (by secondary multiplication of the siphons and tentacles), as well from a phylogenetical 

 as from an ontogenetical point of view (compare above, p. 315). 



Nectosome (PI. XXII. fig. 6, p; fig. 7, longitudinal section ; fig. 8, transverse section). — 

 The pneumatophore has in the Epibulidse the same structure as in the Salacidse and 

 Rhizophysidse ; it includes eight radial bunches of hypocystic villi. Its apex bears the 

 stigma typical of all Cystonectse. The living Epibulia, when it wishes to sink down, 

 expels at will the gas through this apical pore ; as I have observed in Ceylon. -The 

 apical part of the pneumatosac is intensely coloured by a pigment-cap (mitra ocellaris, pp). 



