ORGANIZATION OF RENILLA 503 



breaks up into small cavernous spaces that pass over ultimately 

 into the central cavities of the autozooids and thus communicate 

 through the mouths of these individuals with the exterior. 



In the rachis and the proximal part of the peduncle the two 

 systems of canals, the superior and the inferior, are completely 

 distinct and fluids injected into one system never find their way 

 naturally into the Other. This is not true of the distal part of 

 the peduncle. Here an injection driven distally into one canal 

 flows out freely from the other, showing that the two canals have 

 ready means of communication. The passage from one canal to 

 the other seems to be dependent upon one or more pores in the 

 transverse septum. I found no evidence of a terminal pore con- 

 necting the interior of the peduncle of Renilla with the outer 

 sea-water as described originally by Miiller ('64, p. 354) and 

 more recently for other pennatulids by Musgrave ('09). 



From this description of the anatomical relations of the canals 

 and other cavities in the body of Renilla and from the observed 

 activities of this pennatulid, it is clear that the peduncle is a 

 highly differentiated structure connected with the inflation of 

 the colony. In Renilla the peduncle is not used so generally for 

 burrowing as in many other pennatulids, nor is there reason to 

 suppose that it is especially concerned with locomotion. The 

 fact that a specialized orifice (fig. 1, 0) gn the superior surface of 

 the rachis (R) in Renilla leads directly into the superior canal 

 (/S), which near the distal end of the peduncle (P) communicates 

 with the inferior canal (7) and this in turn opens out through 

 the mouths of the autozooids (Z), is suflEicient when taken in 

 connection with the peristalsis of the peduncle to suggest that 

 this is the system primarily concerned with the inflation of the 

 colony and consequently with its internal currents. In what 

 direction these currents set through the system, however, has 

 never been accurately determined. Agassiz ('50, p. 209) be- 

 lieved that the water entered and left Renilla through the mouths 

 of its autozooids. Miiller ('64, p. 354) stated that the water en- 

 tered through the large central siphonozooid first identified by 

 him. Wilson ('84, p. 725) regarded this opening as the outlet 

 for the system, the water entering through the other siphono- 



