BEHAVIOR 13 



Schaeffer's observations on this point are interesting though 

 negative. As noted, Hve and dead food organisms were not dis- 

 criminated nor were fragments versus whole organisms. When 

 mixed in sugar, beef extract, and other solutions, ink and carmine 

 particles were still predominantly rejected. Likewise, I have 

 observed that the empty but well-formed hulls of long-dead 

 rotifers were quickly rejected at the same time that their live 

 fellows were being eaten. Therefore neither size, shape, taste nor 

 activity of food particles seem to be the basis of selection, which 

 remains a considerable mystery. Schaeffer's results are perhaps 

 the more remarkable because his stock cultures were being fed on 

 something entirely different, viz. ''small paramecia", doubtless 

 with bacteria; and materials like sulphur grains, carmine, and 

 glass particles are probably never encountered by stentors in 

 nature. His demonstration of food selection is rendered more 

 credible by Lund's (1914) evidence of similar discrimination in 

 the related Biirsaria truncatella. 



In a still more closely allied genus, Parafolliculina, Andrews 

 (1947) found that an increase of 10 °C doubles the rate of feeding 

 or the number of food vacuoles formed in a given period of time. 

 Very likely it is the same in stentors, increased feeding being the 

 consequence of thermal acceleration of the membranelles (Sleigh, 



1956). 



It remains, if possible, to locate the site of food selection as we 

 follow the course of digestible material during the feeding act. 

 Particles drawn in by the vortex produced by the membranellar 

 band impinge on the aboral side of the funnel-shaped frontal 

 field and are carried into the buccal pouch by the oralward 

 beating of the rows of frontal cilia. In clouds of carmine, according 

 to Schaeffer, the frontal field cilia may beat circularly, forming 

 balls of this material which then fall over the edge of the disc. 

 If so, this would constitute pre-oral selection, but Dierks (1926a) 

 could not confirm this behaviour; and I found that the feeding 

 vortex simply creates a locus of nearly zero water velocity at the 

 non-oral end of the membranellar band where granules collect 

 until a mass is built up and falls into the rejection current. Current 

 velocity over the frontal field itself is too great to permit such 

 accumulations as Schaeffer described. 



