PHYSIOLOGICAL VARIATIONS IN PARAMECIUM 509 



The difference in conjugating power of different races of Para- 

 mecium has been clearly shown by Jennings: ''The conditions 

 determining conjugation differ greatly in different races of Para- 

 mecium (aurelia or caudatum). Some races conjugate frequentl}^ 

 and under conditions readily supplied in experimentation. Others 

 under the same conditions, conjugate very rarely or not at all." 

 ('10, p. 298). The latter statement is an expression of the result 

 of careful experimentation and should be contrasted with an 

 earlier statement based on a traditional belief; ''Unicellular 

 animals are essentially free germ cells," (Jennings '09, p. 322), 

 a statement that certainly cannot embrace the rhizopods or the 

 sporozoa where germ cells are formed and the parent cells die. 



What Jennings finds true of races we find to be true of different 

 lines within the same race. Some lines will conjugate whenever 

 the conditions favorable to conjugation are prepared; other lines 

 have never conjugated under such conditions. We have found 

 no race (we have called them 'series') as yet in which the con- 

 jugation power of some one line is not much more developed 

 than in the others (for example, J21, G4, Ml, Q3, H' 1, T' 3, 

 etc.) (See diagrams 3-6.) 



It is possible, of course, that under different conditions the 

 remaining non-conjugating lines might have been induced to con- 

 jugate. Under the usual laboratory conditions however, it is 

 evident that the progency of an ex-conjugant is not a homo- 

 geneous race, but consists of differentiated individuals which give 

 rise to pure lines some of which conjugate, others do not. In 

 other words, some Paramecia are potential germ cells; others 

 apparently are not. 



C. Do some Protozoa die a natural death from old age ? 



To return to Woodruff's race of P. aurelia, we find this im- 

 portant difference between his material and that which Calkins 

 worked with. The Paramecium caudatum which formed the 

 material for the earlier observations was a conjugating line as 

 shown by conjugation tests made from time to time (cf. Calkins 

 '02, '04). In this line, therefore, a normal function — ^conjuga- 



