These results demonstrate that in order to obtain the state of stress of rock at depth it is necessary to make actual measurements. The weight of overlying rock must always be an important factor. Topography may have an effect (particularly at moderate depth from the surface) that eventually may be capable of being analyzed, but in addition it is evident that the past tectonic history of the region and the present state of tectonic activity also may influence the state of stress in a manner which cannot be reliably predicted.

The necessity for actual measurements is supported by results of tests reported from other countries. The United States Bureau of Reclamation made pioneer tests of this kind in a tunnel under Boulder Canyon in 1932 (U. S. Bureau of Reclamation, 1939) and later at Prospect Mountain Tunnel, Colorado, and Gorge Tunnel, Washington (Olsen, 1957). The results are discussed by Terzaghi and Richart (1952, p. 61).

E. de St. Q. Isaacson (1957) mentions the existence of natural stresses in the rock on the Kolar Goldfield, India, considerably greater than could be accounted for by the weight of overlying rock. The most numerous investigations have been made in France in the period since 1949 (Talobre, 1957, 1957). In these the direction of. principal stress seldom was vertical, and there was no general tendency for those which approached verticality to be greatest. In a general way stresses increased with thickness of cover, but there was no strict relationship between them. The state of stress was apparently influenced by the tectonic history of the locality. M. Berthier (1958) states that measurements have been made where the horizontal stresses were considerably greater than the vertical stresses. These are considered to be "fossil" stresses from past orogenic movements and locked in place.

INVESTIGATIONS FOR FINAL LOCATION AND DESIGN

Exploration of the Adopted Site

The investigations leading to the selection of this site have already been outlined. The site was finally adopted following the drilling of four sloping diamond-drill holes ranging from 1032 feet to 2005feet in length from the surface. These covered a block of rock at machine-hall level measuring about 500 feet by 300 feet. The results indicated that a sufficiently large area for the power-station excavations would be free from major defects.

A tentative design for the station was prepared then, for the next stage of the investigations, an exploratory tunnel 8 feet in diameter was driven from the floor of the Tumut Valley a distance of about 1100 feet into this site. From chambers near the end of the tunnel, six diamond-drill holes with total length of 1377 feet were drilled across the proposed machine hall and draft tubes.

Determination of the distribution of the granite and gneiss presented considerable difficulties owing to the presence of several sheets of granite, the locally irregular and in many places gradational nature of the contacts between granite and gneiss, the discontinuous and incomplete surface exposures, and the fact that the drill holes were all sloping and in different directions. This problem was solved by construction of a model on which the drill holes were represented by rods, and the information from surface mapping and the exploratory tunnel was plotted on vertical transparent sheets. This model also permitted ready appreciation of the relationships between the various layouts and the geological structure.