450 ROLLIN D. HOTCHKISS 



factors are to be identified with bacterial genes, there should be unmis- 

 takable parallels between these entities. Genes of the higher organisms 

 and bacterial transforming factors have been compared both from a struc- 

 tural or material point of view and functionally.^^ These biological entities 

 both imply the material existence of specific substances at definite sites in 

 the cell. Thus, just as many known genes can be assembled from known 

 sources into one cell, so several independent transforming abilities can 

 be introduced into the DNA of a single bacterial strain.''^ • ''^ On the other 

 hand, like the alternative "mutant" forms of a given gene, the DNA 

 factors for different antigenic and morphological traits*^ ^^ ^^ ^^ •^'^ appear 

 to exclude or even displace each other.''^''^'^^'^^'°° The exceptions to this 

 last statement*'"*^'^"^ are of a sort and frequency interpretable^"^ in the 

 same way as are exceptions among the gene series. Genes are, as we 

 have seen, associated with DNA protein-containing chromosomes — struc- 

 tures having apparently grossly similar chemical composition along their 

 length at different regions which can determine the most diverse biochemi- 

 cal properties of the cell. Likewise, the 30 different transformations men- 

 tioned in Table II are brought about by agents all found in the DNA 

 fractions from the respective strains, fractions apparently uniform in 

 chemical nature and composition. Each of these agents too, like the genes, 

 is reduplicated whenever the cell which acquires it reproduces. 



(2) Mutations in Bacteria. The functioning of DNA transforming agents, 

 like that of genes, is principally studied by the observation of the mode of 

 introduction of a new or variant property, called a mutation, from one 

 strain into another which did not previously bear it. 



In higher forms, a genie mutation is in part defined as a heritable change 

 in the nuclear determinants which occurs as a discontinuous, discrete 

 event and produces a corresponding change in the cellular properties 

 (phenotype). If it is a nuclear determinant which mutates, the mutation 

 should be inherited as a unit gene in crosses; mere transmissibility from 

 mother to daughter cell does not distinguish between a nuclear mutation 

 and a heritable change originating in the cytoplasm. 



Obviously, in identifying bacterial mutations we are at some disadvan- 

 tage. Objects resembling nuclei have been seen in bacteria, ^"-"^"^ but their 

 chemical nature is not yet even outlined, and there is as yet only slight 



if^i H. Ephrussi -Taylor, Cold Spring Harbor Symposia Quant. Biol. 16, 445 (1951). 

 '"2 G. Knaysi, "Elements of Bacterial Cytology," 2nd ed. Comstock Publishing Co., 



Ithaca, N. Y., 1951. 

 i"' C. F. Robinow, Proc. Roy. Sac. (London) B130, 299 (1942). 

 1"^ R. Tulasne and R. Vendrely, Compt. rend. soc. biol. 141, 674 (1947). 

 '"^ K. A. Bisset, "The Cytology and Life-history of Bacteria." E. and S. Livingstone, 



Ltd., Edinburgh, 1950. 



