II. REPLICATION OF DNA IN CHROMOSOMES 97 



are only between whole chromatids, each chromatid is composed of two 

 unlike subunits (Taylor, 1958a). When all of the evidence is considered, 

 the basis for this "twoness" would appear to be the complementary 

 chains of the DNA double helix with some tertiary structure added 

 (see below). 



However, evidence is accumulating whicli indicates that chromosomal 

 fibrils are double at many stages in the division cycle. For a review of 

 the evidence from electron microscopy see papers by Ris (1957, 1961), 

 Kaufmann et al. (1960), and Pappas and Brandt (19611. In interphase 

 nuclei fibrils about 100 A in diameter are seen in cells of many species. 

 Often these are paired. The pairs of 100 A fibrils could very well repre- 

 sent the two chromatids. However, the 100 A fibrils which appear in 

 spermatids before the chromosomes have been reproduced may still 

 contain two DNA double helices, for according to Ris (1961) these 

 change to 40 A fibrils in most spemi nuclei. Although the decrease in 

 diameter could result from a change in the protein with consequent 

 variations in the pitch of a superhelix of the DNA double helix, it is 

 possible that a significant characteristic of genetic material is being 

 overlooked in a preoccupation with models based on single DNA double 

 helices. The studies of Cavalieri and Rosenberg (1961a, b) on the physical 

 properties of isolated DNA and Hall and Cavalieri's (1961) studies on 

 the mass per unit length of DNA fibrils isolated from rapidly dividing 

 E. coll also indicate that DNA from dividing cells is four-stranded 

 while that from nondividing cells is two-stranded, i.e., is composed of 

 single double helices. Is it possible to reconcile these observations with 

 those which indicate that the two complementary chains of a single 

 DNA double helix are the subunits for replication, mutation, and per- 

 haps recombination? 



A model based on a single DNA double helix which extends through 

 the length of a large chromosome without interruption has some of the 

 features required by our present data, but other necessary features 

 would appear to be missing. One of these is the packing or folding of 

 several centimeters of DNA for manipulation within the dimensions of 

 a cell. The DNA or "chromosome" of phage T4 is more than 50 /x long 

 when fully extended (Chapter III). The chromosome of E. coli would 

 be perhaps 50 times this length (more than 2.5 mm). Some of the largest 

 chromosomes might be more than a meter in length (Taylor, 1957). Even 

 if these prove to be composed of several strands of DNA, they would 

 still be several centimeters in length. The most useful models to visualize 

 the folding and unwinding of such a long piece of DNA are those based 

 on a suggestion by Freese (1958). The essential feature of the model is a 

 DNA double helix with a regular sequence of linkers alternating in the 



