The glass transition has been experimentally detected in basalt as (1) an increase in the aggregate linear thermal expansion coefficient αL, (2) an abrupt change in the temperature dependence of Young's modulus dE/dT, and (3) a change in stress relaxation behavior that effectively separates the T> TG and T < TG creep regimes. Transition temperatures determined by the respective experimental methods are (1) 730° ± 15°C, by expansion dilatometry, (2) 725° ± 10°C, by acoustic spectroscopy, and (3) 725° ± 25°C, by stress relaxation. For olivine tholeiite from Kilauea Iki lava lake, Hawaii, thermal expansivity rises through the glass transition interval, attaining αυ = 39.0 × 10−6 °C−1 at 850°C. For olivine tholeiites with moderate glass contents, the temperature dependence of the Young's modulus changes on either side of the inferred TG of 725°C, and dE/dT= −0.071 kbar °C−1 for T< TG, while dE/dT = − 0.21 kbar °C−1 for T > TG. Collectively, the mechanical results suggest that for Hawaiian olivine tholeiite at 1‐atm pressure, the principal material responses are (1) elastic (T ≤ 600°C), (2) reduced creep (600 < T < 725°C), (3) glass transition (T ≃ 725°C), (4) enhanced creep (725 < T< 980°C), and (5) partial melt (T > 980°C). Fracture surface morphologies developed during solidification suggest that the presence of the supercooled melt grain boundary phase may participate in the regulation of the thermal cracking process. Well‐preserved fracture surfaces formed by incremental crack growth are found to be covered with striations that correspond to the inferred sequential stopping positions of the advancing fracture front. These striae may be rationalized in terms of experimental analogues that have been produced in viscoelastic polymers by cyclic tension‐tension loading. The fracture surface morphology, mechanical data, and the controlled crack growth analogues suggest that thermal fracture in solidifying basalt is an incremental and cyclic process, involving three steps: (1) the accumulation of elastic strain energy in cooling rock at temperatures below that required for stress relaxation due to viscous flow in the intercrystalline liquid phase, (2) fracture at a ΔT determined primarily by the aggregate thermal expansion coefficient αυ and Young's modulus E, (3) the penetration by the advancing crack tip, of the thermal horizon capable of relaxing stress due to the creep of intercrystalline supercooled melt, producing the rough surface texture associated with the termination of a striation. Further crack growth must now await the migration of the solidus. The cycle then repeats. Striations measured in deep Hawaiian lava lakes have been compared with the crack advance increments expected in the vicinity of the glass transition, based on two tests: (1) thermal gradients measured in Kilauea Iki, combined with the mechanical properties of olivine tholeiite evaluated near TG, and (2) the crack advance required to match the recorded seismic stopping phases for prexisting cracks of the dimensions expected for Kilauea Iki. The observed versus predicted comparisons are (1) 31 versus 36 cm; and (2) 31 versus 30 cm. We envision this incremental crack growth process as contributing to the control on the downward movement of the thermal cracking front—and its associated hydrothermal circulation zone—in the upper portions of solidifying subaerial and submarine ponded basalt.