Electoral processes provide the central mechanism to the formation of almost all modern democratic governments. Governments operate within a political system. A political system is a part of a larger social system. Social system represents a pattern of human interaction, that bas spatial dimension and operates within a particular situation, milicu or environment. Social characteristics determine political choices. This calls an attention to the study of relationships between the two contemporary systems the 'social systems' and the political systems a subject much discussed by the sociologists and more increasingly by the political scientists, but at the outset much neglected by the geographers.

The entire orientation of our discipline now a days is towards the studies and rescarches regarding the welfare of the societies. All modern governments are claiming themselves to be welfare State. The geographers are also not lagging behind. They have already adopted welfare approaches. The reframing and restructuring of all modern States symbolizing the macro-regions of the world or not due to natural resources but the decision makers who have adopted such decisions through elections to build and rebuild their nations. "When the nation has needed a leader with charisma, it has found or invented one. When the charismatic leader has erred, it has chastised him or her. When leaders and institutions have failed to mediate, the people have mediated through the ballot box and changed the entire landscape." (The Times of India, Jan. 30, 1983, Sunday Review, 1).

The present study aims to examine the comparative characteristics of the electoral response surfaces of sixth and seventh Lok Sabha Elections as influenced by the geographical factors. This has been chosen mainly due to the emergence of the representative areas as the micro-regions of all economic, social as well as political developments. This provides not only new facets to the discipline but an unlimited research frontier. With the framework of the present study a set of 3 major hypotheses have been developed and examined. 1. The electoral participation has direct spatial co-variation with socio-economic determinants of voting behaviour.

2. There is a great regional variation in the spatial pattern of party support and performances.

3. The spatial pattern of party support has been determined by the geographical conditions of the discrete regions.

The entire work is based on the unit of parliamentary constituencies. There is a problem of non-uniformity of collecting units, since the election results are published on the basis of parliamentary constituencies, while the data of socio-economic variables are available on the district level. This has been done away with using interpolation of isopleths by mean elevation method of Robinson and Sale (1961, 10).