This paper deals with a quasistatic, leading to a quasidynamic at a later time, nucleation process that precedes the earthquake dynamic rupture, and models of the earthquake source nucleation are put forward based on physical principles. A quasistatic to quasidynamic rupture nucleation process is an intrinsic part of the ensuing earthquake dynamic rupture; in other words, the nucleation process itself is an short-term (or immediate) precursor that occurs in a localized zone. During the earthquake nucleation, premonitory slip proceeds in the localized nucleation zone, and shear stress also decreases gradually in the zone, since slip-weakening occurs during the nucleation. Immediate foreshock activity is a part of the nucleation process leading to the mainshock dynamic rupture; therefore, hypocentral locations of foreshocks are necessarily restricted to lie near the mainshock hypocentre (onset of the mainshock rupture). Whether or not foreshocks occur during the mainshock nucleation depends on how the rupture growth resistance varies on a local to small scale. Since patches of greater rupture growth resistance are considered to prevail on a local to small scale in the fault zone, carrying immediate foreshocks would be one of the major characteristics of earthquakes that nucleate within the brittle seismogenic layer. When a mainshock earthquake nucleates within the brittle seismogenic layer and its hypocentre is located near the base of the seismogenic layer, immediate foreshocks for this mainshock are necessarily restricted to lie within a localized region shallower than hypocentral depth of the mainshock. By contrast, the nucleation process below the base of the seismogenic layer is aseismic in nature, so that carrying no conspicuous foreshocks would be a common characteristic of interplate earthquakes that nucleate below the base of the brittle seismogenic layer. The breakdown strength tau(p), the breakdown stress drop DELTAtau(b) and the critical slip displacement D(c) are indicative of the rupture growth resistance. To estimate depth variations of these parameters, the effects of the normal stress sigma(n) and temperature T on those parameters are examined using available data so far published. The combined effects of sigma(n) and T on tau(p) in the brittle to semibrittle regime are found to be represented empirically by: tau(p)(sigma(n), T) = tau(p0)(sigma(n))[1 - (cosh(50/T) + 40 sinh(50/T)) exp(- 2000/T)] where tau(p0)(sigma(n)) = 135.7 + 0.750sigma(n), and T is measured in degrees-K and sigma(n) in MPa. It is found that D(c) increases sharply with T, but is insensitive to sigma(n) above 300-degrees-C, while D(c) depends on sigma(n) but is insensitive to T below 300-degrees-C. On the basis of these results, variations in tau(p), D(c) and DELTAtau(b) at mid-crustal depths are estimated for quartzo-feldspathic rocks for a given geothermal gradient.