REPORT ON THE SIPHONOPHOR^E. 257 



pericystic cavity into the same number of radial pouches. Clans (74, p. 22) and Korotneff 

 (50, p. 272) have described peculiar csecal canals in the wall of the radial septa ; but 

 these hypothetical canals are in fact solid fibres composed of exoderm cells, and arising 

 from the pneumadenia (or the air-secreting exoderm of the infundibulum), as has been 

 shown by Chun (48, p. 514). The apex of the pneumatophore has no opening ; but there 

 is a stigma, or a constant opening through which the air is emitted, at the base of the 

 pneumatophore, in the median line of its dorsal side, opposite to the ventral buds of 

 the youngest nectophores ; its place corresponds to that point where, in the Auronectae, 

 the aurophore' is situated. Keferstein and Ehlers (33, p. 3) have already described the 

 spontaneous emission of air by this stigma, and I have repeated the same observation in 

 the Canarian Physophora magnified (84, p. 35, pi. iii. fig. 26). 



Nectophores (PI. XIX. figs. 1-4). — The nectocalyces are in general very similar in 

 form to those of the Agalmidaj. The bilaterally symmetrical umbrella is attached to the 

 trunk by a lamellar triangular pedicle, which arises in the middle line of its concave 

 ventral side. The opposite dorsal side is more or less convex. The principal axis is 

 directed obliquely from above and within, downwards and outwards, so that the apical pole 

 is situated more highly than the basal pole with the ostium. This latter is oblicjuely 

 truncate, often with a pah* of lobiform apophyses on the ventral side. The axial or 

 apical half of the nectophore is always much broader than the basal half, and provided 

 with a pair of auricles or apical horns which embrace the stem. Correspondingly, the 

 large nectosac, which is not much smaller than the surrounding umbrella, is cordate and 

 composed of three parts, a smaller odd basal part and a pah- of large lateral lobes ; the 

 former is subcylmdrieal or subcorneal, the latter are subovate or trapezoidal. The four 

 radial vessels are of very different shape, since the two paired lateral canals enter into 

 the two dilated auricles ; they form here several loops, and are therefore much longer 

 than the two sagittal canals (shorter ventral and longer dorsal) which run simply 

 curved in the median plane of the nectosac. The circular canal which connects the 

 four radial canals above the insertion of the velum is small and ovate, corresponding 

 to the small ostium of the nectosac. (Compare 10, Taf. xxx. figs. 33—35 ; 34, Taf. 

 xxv. figs. 6-8 ; 27, Heft iii. Tab. vi. figs. 1-4, &c.) 



Siphosome (PI. XX. figs. 9-13).- — The trunk of the siphosome in the Discolabidae 

 was regarded by all former observers as a simple sac-shaped and inflated dilatation of the 

 blind basal part of the trunk of the nectosome, and the general opinion was, that the 

 different appendages, beyond the corona of tasters, were more or less irregularly crowded 

 at its basal face (with loose cormidia). This error was not corrected until the year 1877, 

 when the excellent figures of Pliysophora borealis by M. Sars were published in the 

 third part of the Fauna littoralis Norvegise (27, Heft iii. Taf. v. figs. 1-6). This celebrated 

 observer had discovered, many years before, that the siphosome of Physophora is expanded 

 subhorizontally, beyond the nectosome, in form of a large, reniform, spirally-twisted sac, 



(zool. CHALL. EXP. — paet lxxvii. — 1888.) Hhhli 33 



