When was the concept of an "air mass" proposed?
Not long after the electric telegraph made simultaneous (i.e. 'synoptic') observations possible in near 'real time', it was realised that in regions of 'disturbed' weather, (i.e. close to what we now call a depression), two different 'streams' of air could often be found converging into the disturbed zone - each having markedly different properties.
In the British Isles, Robert FitzRoy, the first director of the Meteorological Office is usually credited with highlighting this fact in 1863, though other workers, particularly in France, Germany, Holland and the United States were thinking along the same lines at the same time. Upon the death of FitzRoy, the concept tended to falter, until later workers took up the theme and elaborated upon it: Abercromby in 1887, Napier Shaw and Lempfert in 1911 and of course by the 'Bergen school': V and J Bjerknes and H. Solberg and others during and just after the Great War.
These latter workers proposed the now familiar 'Norwegian model' of the life-cycle of a mid-latitude depression, whereby a minor wave develops along the boundary between two well defined air masses, amplifies (develops) and is carried forward in the general flow. The poleward air mass has an east-to-west component of air motion at low levels, is relatively cold (ex. Polar), and therefore dense, and has a relatively lower humidity value (lower dewpoint) than the 'opposing' air mass. This latter has a generally west-to-east component of motion (at all levels in the troposphere), is warmer (ex. sub-Tropical) and therefore lighter, and has a higher humidity/dew point value. The colder air mass was designated Polar Maritime, and the warmer air mass Tropical Maritime. The boundary between the two air masses came to be known as the Polar Front (see also here and here for other FAQs in this area).