Measurements have been made of (1) light scattering, (2) sedimentation characteristics, and (3) viscosities of identical, well-clarified solutions of a considerable number of preparations of DNA from various sources and under conditions which permit extrapolation of the relevant molecular quantities to zero concentration. In order to increase the range of sizes examined, similar observations have been made with specimens which had been irradiated with 15 MeV electrons or X-rays. It is found that, to a rough approximation, the relation
S
0
α
[
η
]
⅓
holds for all samples (except those which were heated). There does not appear to be any correlation between
M
L
, the molecular weight determined by light scattering and either [
η
] or
S
0
. Values of
M
η, s
from viscosity and sedimentation data have been calculated using the equation of Mandelkern, Flory, Krigbaum & Scheraga (1952) and it was found that with the unirradiated samples these values did not agree with
M
L
, although reasonable agreement was obtained with the irradiated samples. The validity of the calculation of
M
η, s
is discussed. Analysis of the angular variation of the light scattering shows no significant change in the flexibility of the coils upon irradiation, in contrast with the effect of heating where a large change of flexibility occurs. This is in agreement with the hydrodynamic relations given above. Consequently, the differences between irradiated and unirradiated DNA cannot be ascribed to changes of flexibility and the need to use a different model on that account. Since
S
0
and [
η
] are both decreased by small doses of radiation, it is difficult to avoid the conclusion that real changes of molecular weight occur, which have not been discovered by the light-scattering method as employed here. It is shown that both for rods and coils there is an upper limit, at finite angles, to the light scattered by a given concentration of polymer, as the size is increased. For a molecule intermediate between a rod and a coil, the limiting curve may appear to have a finite intercept owing to a concealed high curvature at very small angles of scattering. This may set an upper limit to the observed molecular weight for a given degree of flexibility. If the main effect of irradiation is a decrease of molecular weight without marked changes of flexibility, it is difficult to avoid the conclusion either that single chain breaks do not influence the flexibility of the particle or that a high proportion of double breaks occurs.
