CYTOPLASMIC INHERITANCE 705 



that in plants and Protozoa there are also determinant particles 

 or plasmagenes in the cytoplasm. One of the best known, called 

 kappa, is associated with a nuclear gene K in Paramecium 

 to produce the ' killer ' property ; individuals possessing this 

 liberate into the water something which kills other ' sensitive ' 

 Paramecia. Both K and kappa must be present to produce killer, 

 and both can persist unchanged for generations (K indefinitely, 

 whatever the cytoplasm which surrounds it, but kappa only for 

 four or five cell divisions in the absence of K), and then become 

 active when they meet each other. Plasmagenes do not show 

 Mendelian segregation, but they do show dilution ; that is, their 

 effects bear some relation to the amount of cytoplasm which 

 is inherited. They are self-propagating, and in normal circum- 

 stances their multiplication keeps pace with that of the cell. 

 Sometimes, as by an increased food supply to Paramecium, the 

 cell may be made to divide more rapidly than the plasmagenes ; 

 kappa has b}^ these means been made so dilute (that is, there are 

 so few particles in the cell) that it has no effect, but by a slowing 

 down of the rate of cell division it was enabled to catch up and 

 become active again. 



There is every reason to think that sooner or later plasmagenes 

 will be discovered in the Metazoa. They may indeed be the basis 

 of the dauermodificationen which have been produced in Droso- 

 phila by heat treatment ; these are changes which are inherited, 

 but for a few generations only. Possibly the shock reduces the 

 concentration of, without entirely destroying, a plasmagene, 

 which gradually recovers in the course of sexual reproduction. 



