changes in these properties on freezing.We do not know how the NMR relaxation data in frozen tissue relate to NMR relaxation in unfrozen tissue; we do not know the mechanism of freezing or the alterations in structure and segregation caused by freezing. Thus, we feel neither competent nor obliged to respond to points 2, 3, and 4 above.In their fifth criticism, Chang and Woessner correctly point out (i) that our model neglected the effect of dipolar interaction of the protons in the irrotationally bound water molecules with the protons in the macromolecular structure (8), and (ii) that inclusion of this effect might bring the model into better agreement with the data in the isotope dilution experiment (2). This has also been pointed out to us by Edzes and Samulski (5), who included this intermolecular contribution to T2 for water molecules in the bound state in fitting our proton T2 data for barnacle muscle, extracted its value, and found that it almost dominates the dependence on deuterium concentration in a manner consistent with rapid exchange. They pointed out that it is only the closeness of the deuterium and proton T2's which requires the introduction of the effects of Tb and allows determination of the free parameters in the IBW model. Inclusion of the intermolecular contribution does not disturb the validity of the "incipient motional narrowing" consistency check (3) referred to above.Edzes and Samulski reworked the barnacle T2 data because they were able to derive, from cross-relaxation effects in the relaxation time T1 (albeit of water in chicken muscle), an independent estimate of the intermolecular contribution to T2 in the bound state; the latter agrees very well with that found from the isotope dilution measurements in barnacle muscle and suggests that the model they (and Chang and Woessner) propose for isotopic dilution is valid. The cross-relaxation effects also require (5) that water molecules be bound for times Tb greater than a Larmor period (that is, Tb > 10-8 second); the parameters obtained from barnacle muscle (2,3,5) are consistent with this requirement (Tb -10-5 second). Inclusion of intermolecular contributions to T2 for the protons of water molecules in the bound state does not therefore vitiate the IBW model, but rather sustains it.It was pointed out (1-3) that the IBW model does not hold for protein solutions (9) and agar gels (10). We believe that these two systems may be sufficiently different from muscle tissue that the same theoretical model should not be required to explain the NMR properties of 1182 all of them. The rigid substrate (rigid for, say, tens of microseconds) required for the IBW model may be present in muscle and not in agar gels or protein solutions; it is, in fact, the purpose of the NMR experiments to ascertain these things. With regard to the T1 dispersion data of Held et al. (11), we find support rather than contradiction of the model in question. As far as T10 and T1 dispersion effects are concerned, the muscle systems are, according to the model, effectively ...