Several of the national laboratories have, at some time or other, set up “fundamental” or “primary” standard barometers, with the object of measuring atmospheric pressure in absolute value, and with greater precision than is possible with a barometer of the Fortin type. As long ago as 1854, Kew Observatory established a standard barometer which was described by Welsh. This instrument had a glass tube about 1.1 inches in internal diameter and the height of the mercury column was observed by means of a cathetometer, which was attached to the same supporting wall as the barometer tube itself. The level of the mercury in the cistern was observed by setting the cathetometer microscope on a fiducial mark which was carried on a short steel stem, the lower end of which was ground to a rounded point which could be brought into contact with the surface of the mercury in the cistern. The upper surface of the mercury was observed by setting a horizontal cross line in the cathetometer microscope tangential to the image of the mercury meniscus as seen in the microscope. It is evident that, except for the reduction of the errors due to the variations of the surface tension of mercury, by the provision of a tube of fairly large diameter, this instrument has most of the possibilities of error which may exist in the Fortin type of barometer. The Bureau International des Poids et Mesures had, in 1880, a “normal” barometer constructed on lines suggested by Wild. This intrument was superseded by another, described in the Travaux et Memoires of the Bureau, and which was of interest because of the provision of a simple method of displacing any gases or vapours which might accumulate in time above the surface of the mercury in the closed limb. In each of these instruments the mercury meniscus was observed directly through the barometer tube by means of a cathetometer and the temperature of the mercury column determined by means of thermometers immersed in mercury near the barometer. Subsequently, another “normal” barometer, also constructed on the lines suggested by Wild, was set up. This instrument was of interest in that a new method of observing the mercury surfaces, described by Marek, was employed. This method, which consists in projecting an image of a horizontal wire into the space above the mercury, and reading by means of a microscope the positions of this image and of its reflection in the mercury surface, appears to be entirely satisfactory, and was adopted, with slight modifications, for the primary standard barometer described in this paper.