90 • PROCEEDINGS OF THE AMERICAN ACADEMY 



the surface on which it is creeping, and then contracting the fore part 

 of the body and dragging along the back part enclosed within the tube. 

 Frequently the motion is aided by e version of the pharynx, so as to 

 form a disc or sucker which adheres to surfaces like that of a leech. 

 The movements occur in quick succession, so that the worm crieps 

 about quite actively. At times it doubles itself, — thus passing 

 through its tube and reversing its direction. At times, too, it will 

 leave its tube and creep about without one. The papilhc of the anal 

 aperture are clolhed with vibratile cils, which produce au active cur- 

 rent inwardly as observed in Dero." 



The animals are found either single, or composed of 2 to 4 zooids 

 joined by bud-zones. In March, material kept over winter showed 

 few specimens with bud-zones, and in the great majority of cases the 

 number of segments was 2"). The body tapers slightly posteriorly. 

 It ends in an expanded pavilion resembling that described by Perrier 

 for Dcro oblusa. This pavilion (PL I. fig. 1, poi'.) opens somewhat 

 dorsally, and. when fully expanded, shows no trace of lobes on its dor- 

 sal and lateral borders. On its ventral border it presents in all cases 

 two lobes, which are separated by a median notch and project slightly 

 beyond the rest of the border of the pavilion. When the latter con- 

 tracts, the dorsal and lateral borders become divided into four addi- 

 tional lobes by a median dorsal and two lateral notches. Thus, unless 

 it is examined when fully expanded, the pavilion presents the appear- 

 ance of '"half a dozen prominent, blinit, conical papillie " surrounding 

 the anus, as described by Leidy. When the animal is in plent}'^ of 

 water, the posterior part of the body is habitually kept bent upward, so 

 that the opening of the pavilion is toward the surface of the water. The 

 digit-like appendaiies are attached laterally and ventrally, outside the 

 border of the 2)'ivilion. Their outer ends are slightly swollen. Tiiey 

 project directly backward parallel to one another wiien the pavilion is 

 contracted ; when it is expanded, they diverge. The number of po<laI 

 stylets in each of the fascicles of the four anterior ventral pairs varies 

 from 8 to 14, but they do not differ from one another at all, except in 

 length, and this, taken in connection with the fact that they are fewer 

 and more uniform in size in immature zooids, gives no support to tlh; 

 notion advanced by Perrier ('72, p. G8) for Ue7-o obtitsa, that the fas- 

 cicles are formed by the fusion of ventral and dorsal fascicles. The 

 stylets ill the ventral fascicles differ from those in the dorsal, and one 

 would hardly expect this difference to be obliterated by a fusion of the 

 two. The stylets in these four anterior pairs of ventral fascicles (PI. I. 

 fig. 2) are longer and straighter than those in the succeeding ventral 



