Using the sensitive and stable Setaram microcalorimeter C80, the melting and crystallization traces of ultrahigh molecular weight polyethylene GUR have been obtained at low rates of heating/cooling (upsilon = 0.5, 1, 3, 6, 12 K/h). Slow heating permits identification of three fractions associated respectively with the melting of low molecular weights (I), unconstrained crystals (II), and constrained crystals (III). Melting is arrested between fractions II and III during a time/temperature gap which depends on the thermal history of the sample. Fraction III is found to be highly superheatable, not being able to melt completely below 200-degrees-C if upsilon is higher than 12 K/h. Including H3, the heat of fusion of fraction III, the overall crystallinity is found to be high, respectively 0.87 and 0.94 for the as-received and annealed samples. The equilibrium melting temperatures are about 137 for T(I), 141 for T(II), and above 155-degrees-C for T(III). The high crystallization temperature of fraction III (> 155-degrees-C) reveals that strain in the melt is stable in time/temperature above melting. Heat flows at fusion and crystallization suggest that a hexagonal phase forms between solid and melt in fraction III. The increase of fraction III after slow crystallization links perfection of crystals and strained melting. The possibility that the high-temperature endotherm has another origin than the melting of strained crystals is discussed.