These crystals were on the order of 10 mmalong a side and only about 0.01 mmthick. This was fine, except that X-ray measurements showed that the polymer chains werealigned perpendicular to the large flat faces of the crystals, and it was known that theextended length of the individual chains was on the order of 0.1 mm. How could a chain fitinto a crystal one-tenth of its length? The only answer is that the chain must fold back onitself, as shown in Figure 4.5. Two competing models of chain folding are illustrated: eitherthe polymer folds back neatly on itself or several chains align for short distances withamorphous regions outside the crystal.This folded-chain model has been well substantiated for single polymer crystals. Thelamellae are about 50–60 carbon atoms thick, with about five carbon atoms in a directreentry fold. The atoms in a fold, whether direct or indirect reentry, can never be part of acrystal lattice.It is now well established that similar lamellar crystallites exist in bulk polymer samplescrystallized from polymer melts, although the lamellae may be up to 1 mm thick. Recentresults support the presence of a third, interfacial region between the crystalline lamellaeand the amorphous phase. This interfacial phase can make up about 10–20% of thematerial. Furthermore, there does not seem to be much, if any, direct-reentry folding ofchains in bulk (or melt)-crystallized lamellae. This is illustrated in Figure 4.6. Orientationof the lamellae along with additional orientation and crystallization in the interlamellaramorphous regions, as in the fringed micelle model, is usually invoked to explain theincrease in the degree of crystallinity with drawing (stretching).