where p0 is the ideal/nominal pressure and dT/dp is the pressure coefficient defined by ITS-90 (values are given in Appendix B). For melting and freezing points in open cells without any pressure control or measurement, the pressure is determined by the weather and altitude at the location of the measurement. This leads to pressure variations (standard uncertainty) with weather, over periods of days, of about 2 kPa (2% atm.), and pressure differences with altitude of 20 kPa or more, leading to corrections of 1.5 mK or more. The uncertainty in the pressure correction is given by measmeas0pdTdTuTupuppdpdpwhere u(pmeas) is the uncertainty in the pressure inside the cell and u(dT/dp) is the uncertainty in the pressure coefficient. If the cell is subject to pressure control or measurement, the uncertainty in the pressure may be 10 Pa, or lower. The uncertainty in the pressure coefficient is practically negligible, and ranges from about 0.2% for the triple point of water to a few percent for the other fixed points (McAllan 1982). In open cells, the pressure coefficient is easily confirmed by changing the gas pressure. The current values used for the fixed points are also consistent with values calculated from the Clausius-Clapeyron equation: fpfTVdTdpH∆=∆, (2.3) where ∆Hf is the molar heat of fusion, Tfp is the fixed-point temperature, and ∆V is the molar volume change. With sealed metal fixed-point cells, the internal cell pressure is not measurable, and the operator must rely on the manufacturer of the cell having sealed the cell at the correct pressure. With the usual procedures, and in the absence of leaks, the pressure in the cell when at the fixed-point temperature is approximatel