abutment on the Pacific slope. During its early years of high costs and limited earnings in British Columbia, the CPR concentrated on the development of more remunerative local main- and branch-line traffic rather than through or transcontinental traffic. For the operation of the CPR Pacific slope line, then, Innis implies that success meant simply controlling expenditure and limiting losses that were covered by returns from more profitable sections of the line. ²⁴ For the construction and operation of its lines in the Kootenay region in the southeastern corner of the province, Innis elaborates the CPR's strategy. By establishing a low tariff that encouraged mine owners to ship their ore, the CPR elicited extensive local traffic that reduced the railway company's overhead costs over long stretches of otherwise non-revenue-producing territory and relieved in part the cost of settlement. ²⁵ The most recent overview of railway development in British Columbia sees this ability to generate traffic (or capture it from competitors) as the key to the success of the Canadian Pacific and the failure of both the Canadian Northern and the Grand Trunk Pacific. ²⁶ For Innis, however, the success of the CPR in British Columbia as well as on the prairies also advanced 'the dominance of eastern Canada over western Canada.' ²⁷

The owners of the Canadian Northern, the last transcontinental to enter British Columbia, also attempted to encourage traffic along a section of the line requiring greater construction expenditures and offering much less local traffic than the prairies. In his detailed study of the enterprise, Regehr asserts that the 'development of local traffic resources was certainly very much a part of Mackenzie and Mann's British Columbia plans.' To generate freight, they purchased the Dunsmuir Colliery on Vancouver Island, built a huge sawmill at Fraser Mills near Vancouver, and invested in numerous hardrock mining and fishing enterprises. These activities would, one of the Canadian Northern officers claimed, have produced enough traffic by 1920 to make the British Columbia section pay not only its operating expenses but also its fixed charges. ²⁸

'Not following the CPR example' may embrace the causes of the GTP failure as a unit, but it does not offer explicit measures of failure for the different components of a large, complex organization such as a railway company. Though he concentrates on success in his studies on the evolution of the managerial structure of American corporate enterprise, business historian Alfred D. Chandler, Jr., also suggests functional standards of failure. For senior managers who allocate the resources of the

-10-