128 BULLETIN OF THE 



their nature, and give the early condition of the knob a likeness to that 

 of certain other genera of Physophoridse. 



The knob of Fhysophora hydrostatica originates, like that of other tu- 

 bular Jelly-fishes, as a simple bud, hardly distinguishable from the earlier 

 condition of all structures in the Siphonophores. In its place of origin it 

 resembles the genus Agalma, for it forms on the ciliated base of the feed- 

 ing polyp, and is in fact a proliferation of the walls cf that part. Whether 

 all Physophore knobs originate from the same relative position is an 

 open question. In Rhizophysa filiformis we have several of the polypites 

 nearest the float with naked tentacles, from which the knobs bud, never 

 arising from the base of the polypites. Of course these imdeveloped 

 appendages in the singular Rhizophysa may be looked upon as tasters,* 

 a supposition hardly probable ', or it is also possible that they correspond 

 with somewhat similar structures between the nectocalyces of Apolemia 

 uvaria. In the well-known genus last mentioned the polyp-like parts 

 between the swimming-bells appear to have no filaments like those 

 found on the tasters of other Siphonophores. The taster-like bodies near 

 the float of Rhizophysa are undeveloped feeding polyps. 



In the very earliest stages the Physophora knob is composed of layers 

 which are apparently two in number. The differentiation of other 

 layers takes place later in the course of the development. At first we 

 find only ectoderm and entoderm in the walls of the knob. This simple 

 bud elongates into a flask-shaped body, at the base of which the cavity 

 becomes enlarged, imparting to this region a more or less spherical shape. 

 (PI. I. fig. 2.) From the ectodermic wall of the enlargement thus formed 

 arises the involucrum. An lamination of this region, even at an early 

 period (PL I. fig. 3), shows that a differentiation has begun, and that the 

 ectoderm has divided into two layers, one of which appears as a collar 



* The word taster, to designate peculiar structures among the Siphonophores, is 

 perfectly applicable in the case of Physophora. In other genera the designation 

 " Saftbehalter " may seem better ; but here in Physophora their function seems dif- 

 ferent from that of the same part in most of these animals. The filamentary append- 

 age to the taster in Physophora, although very easy to see, has been overlooked by 

 several naturalists. (See Kolliker, Schwim7npolypen von Messina. Vogt, Siphono- 

 phores de Nice. Leuckart, Siphonophoren von Nizza, p. 106, ) According to Kefer- 

 stein and Ehlers [l. c, p. 31), these appendages to the taster were discovered by 

 Philippi, but were omitted in the descriptions by naturalists who followed him until 

 the investigations of Sars. In his Anatomy of Physophora, Clans (1860) speaks of 

 them (p. 17), but has no representation of the filament in his figure of the genus. 

 (Claus, Ueber Physophora hydrostatica 7iehst Bemerkungcn iiber andere Siphonophoren, 

 Zeitsch. f. Wiss. Zool., Bd. X. p. 1, fig. 1. Philippi, in Miiller's Arch. f. Anat. u. 

 Physiol., p. 61, Taf. 5, fig. 4. Sars, in Middelhavet's Littoral Fauna, p. 4.) 



