300 C. LEVINTHAL 



Most of the alternative models which have been proposed involve the forma- 

 tion of recombinants during the replication of genetic material. If replication 

 occm*s by the laying down of new material along the parental structm-e as a 

 template, then the recombinant can be formed if a partially formed daughter 

 switches to copy along another parental structure before it has completed its 

 growth. The recombinant is then a partial replica of one parent combined 

 with a partial replica of the other, and this type of model has been called 

 partial replica formation or copy-choice (see Fig. 4). 



One of the consequences of this type of model is the maintenance, both 

 chemically and genetically, of the integrity of the parental structure. It is 

 the chemical integrity of the parental structure which the third type of experi- 

 ment on the nature of the mating event has been designed to test. For 

 chemical experiments to be relevant to the genetic analysis, one must make 

 certain assumptions as to the nature and the continuity of the carrier of the 

 genetic information. It is known from the experiments of Hershey and Chase 

 (1952) and Hershey (1955) that the genetic information of the phage enters 

 the host bacterium as DNA (deoxyribonucleic acid) and that it leaves the 

 infected cells m newly formed phage as DNA. However, one camiot be sure 

 that the information does not transfer to some other chemical structure 

 during growth in the infected cell. Since there is no compelling evidence to 

 the contrary (see Stent (1958) for a summary of the opposite view), it will be 

 assumed here that the genetic information remains in DNA and, more 

 specifically, that it is the DNA which is involved in the mating events. Thus, 

 if one grows phage labeled in its DNA, one can ask if the label is distributed 

 to the two products of a mating, as would be expected in a crossing-over 

 model, or whether it remains in one particle, as would be expected in a 

 partial replica-type model. 



Tracer experiments designed to study the way in which phage particles 

 are produced in the infected cells are discussed in detail in Chapter 7. As was 

 pointed out there, two different techniques (Levinthal, 1955, 1956; Stent and 

 Jerne, 1955; Delbriick and Stent, 1957) have led to the same- general con- 

 clusions as to the distribution, among single-phase particles, of the P^^ atoms 

 which are transferred from labeled parents to progeny. Of the average, about 

 half of the P^^ in labeled parental particles is found in the progeny phage after 

 one cycle of growth in unlabeled bacteria. And among the progeny particles 

 there is on the average about one per infected cell which contains approxi- 

 mately 20 % of the P^2 in one of the labeled parents. Or, to describe the results 

 differently, about 40 % of the P^^ which is transferred to the progeny of each 

 bacterium winds up in a single phage. If a second cycle of growth is allowed 

 in unlabeled bacteria, the total amount of P^^ transferred is again about half; 

 thus, a quarter of the phosphorus atoms of the first cycle parents is found in 

 the second cycle progeny. However, the fraction of these transferred atoms 



