MUSEUM OF COMPARATIVE ZOOLOGY. 273 



in part make up for the loss of swimming-bells and propel the animal in 

 the water. As far as my observations go it is rather sluggish, and but poorly 

 adapted for rapid progression. As the colony floats along, the covering-scales 

 are generally extended to their greatest width, but when alarmed the scales 

 are contracted closely together around the feeding-polyps and those other organs 

 which arise near their base. 



The polypites, tastern, and sexual bells arise from a slight enlargement of a 

 structure corresponding to the axis of the other Physophores. This enlarge- 

 ment does not diff'er greatly from a similar sac at the extremity of the stem of 

 the genus Physophora. The polypite (p) resembles the same structures in 

 other Siphonophores. They are long, flask-shaped bodies, projecting beyond 

 the covering-scales (c) when extended, and have a mouth at the free extremity. 

 The walls have a pink color with patches of dark crimson pigment near the 

 base of the feeding-polyp. From each polypite, near its origin, there hangs 

 a single tentacle. This tentacle (t) is dotted along its whole length by sec- 

 ondary appendages or tentacular knobs (A:), of which there are two very differ- 

 ent forms. 



The existence of more than one kind of tentacular pendant in an adult 

 Physophore * has up to this time been thought peculiar to the genus Rhizopihysa, 

 Two forms of these bodies exist in the young of several genera. In the young 

 Agalma we find the permanent knob of the adult coexisting with an embryonic 

 form. The same is true of the young Agalmopsis and Nanomia. Each kind 

 of knob, however, in the larval Physophore, is limited to its own tentacle, and 

 in the case of the embryonic knob the tentacle itself has the same provisional 

 nature as the structure which it carries. 



The adult Rhizophysa f has three difl'erent kinds of tentacular knobs hanging 

 from one tentacle, and as far as we now know none are embryonic. A similar 

 condition exists in A. formosa, with the exception that there are here two forms 

 of tentacular knobs instead of three. 



The first and more numerous kind of tentacular pendant (PI. VI. figs. 9, 10) 

 is in many respects like that of the adult Agalma. It has a sacculus (rf), an 

 involucrum, and two terminal filaments (b), one on either side of a spherical 

 bladder or vesicle (c). The terminal filaments are, however, shorter than those 

 of Agalma, and are commonly carried stiflHy elevated like two horns. 



The peduncle of the knob is very flexible and of moderate length, admitting 

 a free motion of the pendant in all directions. The involucrum closely sur- 

 rounds the sacculus, and its walls are with the greatest difficulty distinguishable 

 from those of the latter body which it contains. The saccalus has a single coil 

 upon itself, and its walls have a dark crimson color. At its base there are 

 rows, generally two in number, of large lasso-cells, homologous to similar 

 bodies in the knob of other Physophores. 



* See reference to Sars's genus Agalmopsis on preceding pages. 



t Proc. Best. Soc. Nat. Hist., XX. These three kinds of knobs in R. filiformis 

 are distinct from their earliest embryonic coudition, and do not develop one from 

 another. 



VOL. IX. — NO. 7. 18 



