Please explain what figure 5 means

Fic. 4. Mean residue-te-revilus aecupaney an a funtetion of peeition in the molecule. The abscisan gires the s co-ontinste of the phosphorus atoes in each raidoe. Bince the long axis of the moleeule in almost perallel to t, thin diegram shows a vinw of the molecule similar to that ahown in Fig. 1(b). The oecupsneice are on an arbitrary scale. Resilae numbers are ahown beside asme of the eireles. Fic. 5. Watcon-Crick base-pairs nuperimposod os mir. denaity. (o) A39-U41 in the settcodon atem. Here the desuity is wak and the base poeitions are in -determined. This region of the map did ant improve subetantially en refineenent. (b) cs3-C61 in the TYC helix, showing atrong density and well-defined bases. CRYSTALLOGRAPHY OF YEAST RNA Regions of the molecule which showed suppicieualy large shifts in any cyele were plotted into the m.i.r. density ss a eheck against systemstie drif auray from their most likely atomie positions. An example of this problem is ahown in Fig. S(a), which shows the base-pair A20-U41 of the anticodon superimpoeed on ita mi.r. denwity. Hiere the bese positions are not well defined by the density, and their movements during real-space refinement are atrongly influenoed by movemonta of the beckbone to which they are astached. For comparison, Fig. 5(b) shows the base-pair G53-C61, which has rather good denaity and is thus well-poaitioned. At the atart of eycle 4 we added 70 cations or water molecules which sppeared as isolated peaks in the mi.r. map. Many of these did not peraiat on refinement, whiles a large number of new pealse appeared. Ultimately we retained anly those eations which perainted from eyele to eyclo, and which wore in stereochemically reasonable poaitiona. Fic. 4. Mean residue-te-revilus aecupaney an a funtetion of peeition in the molecule. The abscisan gires the s co-ontinste of the phosphorus atoes in each raidoe. Bince the long axis of the moleeule in almost perallel to t, thin diegram shows a vinw of the molecule similar to that ahown in Fig. 1(b). The oecupsneice are on an arbitrary scale. Resilae numbers are ahown beside asme of the eireles. Fic. 5. Watcon-Crick base-pairs nuperimposod os mir. denaity. (o) A39-U41 in the settcodon atem. Here the desuity is wak and the base poeitions are in -determined. This region of the map did ant improve subetantially en refineenent. (b) cs3-C61 in the TYC helix, showing atrong density and well-defined bases. CRYSTALLOGRAPHY OF YEAST RNA Regions of the molecule which showed suppicieualy large shifts in any cyele were plotted into the m.i.r. density ss a eheck against systemstie drif auray from their most likely atomie positions. An example of this problem is ahown in Fig. S(a), which shows the base-pair A20-U41 of the anticodon superimpoeed on ita mi.r. denwity. Hiere the bese positions are not well defined by the density, and their movements during real-space refinement are atrongly influenoed by movemonta of the beckbone to which they are astached. For comparison, Fig. 5(b) shows the base-pair G53-C61, which has rather good denaity and is thus well-poaitioned. At the atart of eycle 4 we added 70 cations or water molecules which sppeared as isolated peaks in the mi.r. map. Many of these did not peraiat on refinement, whiles a large number of new pealse appeared. Ultimately we retained anly those eations which perainted from eyele to eyclo, and which wore in stereochemically reasonable poaitiona