49 The data portrayed in the Figures centre about the marginal farm. This is represented by the straight line drawn through zero farm surplus. The marginal farm would represent that farm which, during the year of the study, paid operating and depreciation expenses and $700 to the operator. It netted no surplus for interest on investment, debt retirement or savings. A marginal farm, then, would fall short of being a satisfactory farm unit in at least two respects. First, a farmer with owned capital legitimately could expect to earn a fair rate of interest on it. Second, a farmer with borrowed capital would need a surplus with which to make the annual payments of interest and principal. The latter farmer definitely would require additional income if living expenditures were to be maintained at $700 per year. The satisfactory farm, then, is somewhat larger than a marginal one. The probable surplus also gives an indication of the debt-carrying capacity of the farm. As an example, the operator of a farm which netted a surplus of $100 would be able to make the annual payment necessary to amortize a loan of $1,485 in 25 years at an interest rate of 45 per cent. From the figures the marginal farms appear to be approximately as follows: 19 productive livestock units and 63 acres of cropland at Prince George; 21 productive livestock units and 90 acres of cropland at Vanderhoof; 29 productive livestock units and 135 acres of cropland at Smithers; 37 productive livestock units and 135 acres of cropland at Francois Lake. It is necessary to make some reservations in interpreting the results shown. The crop in the Smithers and Francois Lake districts was a partial failure in the year of the study and it is improbable that it was completely offset by the favourable price of farm products which prevailed. In all probability, then, the farm unit in those districts that will return a positive farm surplus over a period of years is somewhat smaller than is indicated by the data. At Prince George and Vanderhoof the crop was more nearly normal. It may be, then, that because of the favourable price situation prevailing at the time of the study the minimum size that would ensure solvency on a long time basis would be somewhat larger than was indicated by the data. CONCLUSIONS The study indicates that the rate of progress in agricultural development in the area had been fairly slow. This serves to emphasize the difficulties which face a settler who attempts to improve forested land. The fact, too, that farm income per farm in the area was only one-half of the provincial average suggests the desirability of a considerable adjustment in many farm businesses. Two factors which have been instrumental in retarding agricultural development in the area are worthy of consideration in the light of this study. The first is the cost of bringing the forested land under cultivation. The second is the handicap of comparatively high shipping costs from the area. The first factor is particularly significant in view of the direct relationship which appeared to exist between size of farm business and income. Under typical production practices of the area livestock and improved acreage were the important components which determined size of farm business. Improved acreage, though, tended to be the limiting factor for land had been brought under cultivation only at a fairly slow rate. By hand labour it would require 15 to 65 years to clear the 60 to 135 acres needed for the minimum satisfactory farm unit. It was understandable, then, why more than one-half of the farms were still below the minimum satisfactory size.