How deep can be a dyke?

The study of the three dimensionality of a dyke is fraught with many difficulties, yet it is an important variable that needs to be characterized to fully understand processes of lithospheric magma transport. The difficulties encountered when attempting to answer the question of how deep can be a dyke are not limited to the lack of an adequate exposure of dykes in the field, but also include several aspects of the common simplifications introduced in theoretical models aiming to describe the formation of this type of intrusive, and even extend to include aspects of the conceptual models of magma ascent. All of these problems are examined in the present paper, showing how is it that each of them can influence the answer that a particular person gives to such question. Although a revised model of the conditions leading to dyke formation is advanced here, it is shown that the present uncertainty surrounding the state of stress in the Earth's crust, its distribution in space and evolution in time inevitably lead to a wide range of possible depths for a given dyke. Nevertheless, by removing the sources of ambiguity inserted in current models portraying dykes as fluid-filled cracks, much is gained towards providing a more definitive answer to this question. In any case, based in the analyses made here, it would seem that the existence of dykes deep enough to provide a temporal conduit between the deepest magma sources and the surface seems to be favored in most conditions of geological relevance.