Cement pastes were prepared using Cement and Concrete Reference Laboratory (CCRL) proficiency cement samples 140 [9] and 152 [10]. Since many of the experimental procedures have been provided in detail elsewhere [11-13], they will only be reviewed briefly here. Cement and water were mixed in a high speed blender using the following protocol: 30 s of low speed mixing while the cement powder is introduced into the mixing vessel that already contains the water, 30 s of high speed mixing, a rest of 1 50 s while the sides of the mixing vessel are scraped down, and 30 s more of high speed mixing to prepare the final product. The mixing water was either distilled water or a solution of alkalis (sulfates or hydroxides), prepared by dissolving the relevant potassium and sodium compounds in distilled water. The fresh paste was used for Vicat needle penetration [14] measurements at laboratory temperature (about 26 ºC) and rheological measurements by the stress growth technique [15-17] at 20 ºC, or cast into small (≈5 g) wafers that were placed in capped plastic vials and cured under saturated (small amount of water on top) or sealed conditions in an environmental chamber maintained at 20 ºC. Limited repeatability tests have indicated that the calculated values for yield stress have a standard uncertainty of 5 % [15-17]. At various ages, the specimens were removed from their vials, crushed using a mortar and pestle, and analyzed using loss-on-ignition (LOI) to assess the degree of hydration and/or LTC to investigate their pore structure. Based on a propagation of error analysis, the estimated uncertainty in the LOI-calculated degree of hydration was 0.004 [13]. For temperatures between −100 ºC and 500 ºC, the LTC equipment manufacturer has specified a constant calorimetric sensitivity of ± 2.5 % and a root-mean-square baseline noise of 1.5 µW.