Gravity meters measure the difference in gravity between a base station (where the absolute value of gravity is known) and a series of field stations. Most base stations are established by measuring the difference in gravity between the new base and an already established base station. Many base stations trace their origin back (through a line of base stations) to Potsdam, Germany, where physicists attempt to measure "g" to a precision of one part in a million.
I prefer the LaCoste and Romberg meter over others I have used. Instructions for operating the L&R model "G" meter are posted at http://www.seismo.unr.edu/ftp/pub/louie/class/field/lacoste.html
Once the gravity at a new field site is known, theoretical gravity is calculated. Theoretical gravity depends on latitude, elevation, and on the surrounding topography. Some equations can be found at http://horton.lbl.gov/users/cscott/docs/snthesis/field.html although it would be better to consult a good geophysics textbook or take a class in gravity field methods. The difference between the observed and theoretical gravity is called the Bouguer anomaly. If the Bouguer anomaly is negative, it means the observed gravity is less than the theoretical.
Because the force of gravity is proportional to the mass responsible for the gravitational field and inversely proportional to the square of the distance between any part of that mass and the observation point, a local lack of normal mass (say, a thick layer of low-density sediments instead of heavy igneous rocks of negligible porosity) will result in a local gravity low. A mass of unusually dense rock (gabbro intruding sedimentary rock or acidic volcanics) will generate a gravity high.
A case history is in preparation.