130 THE BIOLOGY OF STENTOR 



habit in emphasizing major issues in the remainder of the paper. 

 Hence it seems to me that all later commentators have mis- 

 interpreted this passage as a statement that membranelle formation 

 can occur in place or without primordium formation. But this is 

 not to exclude that such development may occur after a fashion, 

 anomalous as this w^ould be. Schw^artz repeated Prowazek's salt- 

 shedding experiments and stated that in some cases there W2is 

 clearly a neo-formation of membranelles in situ. His explanation 

 was that only the cilia of the membranelles had been cast off, 

 leaving the basal bodies intact, from which new cilia may have 

 grown ; and he remarked that if this can occur, such replacement, 

 rather than primordium formation, should be the method of 

 renewing supposedly worn-out feeding organelles. I have myself 

 noticed a few similar cases. In one of these, a stage-3 divider was 

 treated with sucrose and it shed the membranellar band. The 

 division primordium remained but showed abortive development, 

 while around the anterior rim of the cell there appeared within 

 about 4 hours shorter than normal oral cilia which beat in meta- 

 chronal rhythm. Such cases indicate that if carefully graded treat- 

 ments were employed, a renewal of the large oral cilia if not the 

 entire membranelle might be firmly established. Yet it is certain 

 that in most experiments of this sort the entire band comes off and 

 the regeneration primordium is soon formed (see p. 252). 



8. Repeated oral regeneration 



Since the formation and development of an oral primordium 

 involves the production of thousands of new, large, oral cilia as 

 well as other parts, one wonders whether there is an inexhaustible 

 reserve for such synthesis. Gruber (1885b) cut and presumably 

 decapitated a coeruleus on 5 successive days and each time 

 complete regeneration followed until the animal finally became 

 necrotic and too small for further operation. With the same large 

 species, Prowazek (1904) also performed successive cuttings. In 

 one tabulated case an animal was cut nine times during which 

 macronuclear beads were not removed, and this animal always 

 regenerated. The material of the macronucleus seemed to have 

 been substantially drawn upon, because it was finally reduced 

 from II to only 2 nodes. He also reported 3 cases in which the 

 animal was repeatedly cut or wounded and compelled to re- 



