Detailed weather station histories are vital for identifying causes of anomalies in the observational data. Changes in the observation site's location, elevation, exposure, and instrumentation are known to cause such anomalies. The station histories do not contain a record of the changes in the criteria that were used to initially position an observation site or those subsequently used to move it. Neither are they accompanied by the philosophy from which those criteria were derived. Both the philosophy and the site criteria have changed significantly over the years. The first national observational network was that of the U.S. Army's Surgeon General in 1817. That network followed the expansion of the settlement frontier with observations at the Army posts and provided the first weather information recorded in most states. The Smithsonian Institution's network recorded weather data from about 1847 to 1871 and collected them expressly for climatic research. That changed when the U.S. Army's Signal Service network was created in 1870 to become a weather network that used telegraph to collect data, prepare surface maps, and distribute forecasts. The Weather Bureau morphed from them into a network that supported agriculture's needs. Its transfer from the Department of Agriculture to the Department of Commerce was followed by relocation of weather observations to airports. It was continued by the National Weather Service's network whose primary focus remains on weather rather than climate. Each of the national networks had criteria for site selection and each had a philosophy from which those criteria emanated. This paper presents a brief discussion of those philosophies, how they differed from one network to another, and how they influenced their site selection criteria. Examples of site selection criteria will be presented. Those used by the Signal Service and the Weather Bureau from 1870 to the 1940's dictated that the tops of the high buildings in the middle of cities met the requirements of preferred locations. The current views of a what constitutes a good weather observation site will be compared to those of the past.