268 CELL HEREDITY 



can multiply only there, or perhaps because it is integrated and there- 

 fore noninfectious in other tissues. The great difference in susceptibility 

 between XX and A'V tissues may also be interpreted in terms of different 

 rates of replication or degrees of integration of the agent with the host. 

 The infectious sex-ratio agent has just recently been described, and 

 future investigations may be expected to provide new evidence on these 

 key questions of replication and integration of an infectious determinant 

 with the hereditary material of its host. Chromosomal genes which in- 

 fluence sex ratio are also known and have been extensively investigated. 



CONCLUDING REMARKS 



In this chapter we have considered two problems: the demonstration 

 that nonchromosomal determinants exist, and the further question of 

 their properties. The principal criterion we have chosen for identify- 

 ing the class of nonchromosomal determinants is the inability to locate 

 the factor on a chromosome. Applying this criterion has netted us a 

 heterogeneous collection of systems which require further analysis. 

 Nonmapability is a safely inclusive criterion for initial selection because 

 any factors which do map are chromosomal. Further supporting evi- 

 dence adduced has varied from system to system, and includes: (1) per- 

 sistence and stability of the phenotype under conditions of cell reorgani- 

 zation (fertilization and meiosis) and environmental change; (2) mutabil- 

 ity, especially a very high rate of mutation induced by a particular 

 chemical; (3) infection in the sense of cytoplasmic spread; (4) uniparental 

 transmission; and (5) somatic segregation not attributable to chromo- 

 somal events. 



Approximately 100 well-characterized instances have been reported of 

 determinants which meet one or more of these criteria, but from the 

 existing information, it is fairly evident that they include different sorts 

 of systems. In an attempt to organize this material into categories which 

 might be useful for further analysis, we have considered four hypotheses 

 proposed by different investigators at different times, concerning the 

 nature of these systems. They are not mutually exclusive; it is even 

 likely that examples of all four types may coexist within an organism 

 simultaneously. 



One property shared by several nonchromosomal determinants is their 

 high rate of mutation or loss following treatment with particular chemical 

 mutagens. The acridine dyes, such as euflavin, trypaflavin, and acridine 



