VihridUc Tufis or Flame Cell in Rotifera. 131 



in ;ill tlie tiguics given, was attached what appeared as a solid 

 and, as compared with the walls on either side, a thick flagellum, 

 extending nearly to the point of attachment of the tag to the 

 lateral canal, and down this apparent flagellum ran a series of 

 undulations connnencing at the attached end of the movable 

 body and passing along to its free end. 



Those in tlie flat aspect showed the plug at the distal end as a 

 border of protoplasm running across ; it was spongy looking, 

 occasionally vesiculated, and with little protuberances on the 

 outside, generally looking denser just at the point of attachment 

 to the undulating body, and in appearance analogous to the 

 suV)stance of the walls of the lateral canals. As in the narrow 

 aspect, the sides were excessively thin. Indeed in a dead 

 animal the whole appearance was that of a hyaline fan-shaped 

 cavity, the boundaries of which required most careful focus- 

 sing to define. In a living and vigorous animal, the flickering 

 appearance which has given rise to the term " flame cell," was 

 very rapid, and it was difficult to determine the exact nature of 

 the movement going on, but in animals treated with cocaine, or 

 losing vigor through long confinement, the movement could be 

 distinctly seen as a series of waves in some substance lying 

 between the upper and lower surface of the tag. Through this 

 moving substance could be focussed, and kept steadily in view, 

 two distinctly longitudinally striated surfaces. Repeated obser- 

 vations on cells in various states of activity, confirmed the 

 conclusion that these striations are on the walls of the tag and 

 not on the undulating body. Further, the most careful scrutiny 

 failed to show any lateral borders to the moving body other than 

 the sides of the tag itself. 



In a recently killed animal the undulatory movement could be 

 seen to die down until it was a slight wave-like appearance, 

 gradually narrowing until it occupied the median portion of the 

 tag, and dying out before passing more than half way down, 

 just as a ripple on water dies away as it recedes from the point 

 of disturbance. Another appearance obtained from some chance 

 views of a tag when the free end was pointing up the microscope, 

 is shewn in Fig. 3. Here was presented an optical section 

 showing a flattened oval with thin walls and a thicker line 

 running between and joining the extremities. 



