T here is nothing original in the proposition that mechanical stress has a controlling influence on the propagation of electrical trees in insulating materials. Over a quarter of a century ago, Billing and Groves [1] and later Arbab and Auckland [2] were discussing the effect of externally applied mechanical stresses-for example by bending-on tree growth. In a simple bending experiment, it is immediately obvious that tree growth is accelerated in regions of tensile stress and retarded where compressive stress is present. The effect of internal stresses on electrical tree growth, determined by photo-elastic techniques, was also reported by Champion, et al. [3] and David, et al. [4] in 1992 and 1994, respectively. Whereas the above were concerned with laboratory specimens, similar conclusions have also been drawn from more practical applications, such as in cables, in numerous publications from Densley in 1979 [5] through to Ildstad and Hagen in 1992 [6].The theoretical analysis of the mechanical aspects of tree growth began with Zeller, et al. [7], [8], who attributed the growth of filamentary cracks to electrostatic forces, a process they named electrofracture. This theoretical approach has been developed subsequently by Hikata, et al. [9] and Fothergill [10].On the basis of this approach to an understanding of the physical processes, there is a strong temptation to interpret electrical tree growth in terms of fracture mechanics, but the analogy must not be pushed too far. Certainly, it is possible to draw a direct correlation between treeing characteristics and the mechanical properties of insulating materials. A preliminary study of tree inception and growth in materials of greatly differing mechanical properties [11] revealed that the mechanical characteristics of the material were paramount. It was proposed that the residual mechanical stress developed under alternating voltage application produced fatigue cracking in the inception phase. Since crack formation is determined by, among other things, tensile strength, it followed that this parameter would influence the tree initiation time. This view was supported by measurements of the tree inception times of polyester resin and an elastomer (PL-3). The tree inception time in the elastomer was less than 30% that of the polyester resin, which had a tensile strength five times greater than that of the elastomer.The buildup of internal tensile stress due to the alternating electrostatic forces, as well as leading to crack formation and tree initiation, will also increase the material's propensity for tree growth. It follows that tree growth in elastic materials may be less rapid than in inelastic materials, since the former will not suffer from long-term mechanical stress buildup. Thus, a material's modulus of elasticity could provide a measure of its resistance to tree growth. This was confirmed by a comparison between PL-3 and polyester resin, which demonstrated an inverse dependence of the time for a tree to cross the interelectrode spacing on the modulus of...