L\—l.

trains per annum, from Table XL, and by 312 full working days in a year, gives the annual cost of assistant locomotives standing still in pounds sterling. These results are shown in Table XXII. Adding together the annual cost of fuel from Table XII., of wages, renewals, and repairs, and supplies from Table XXI., and the cost of assistant locomotives standing still from Table XXII., gives the total annual cost of motive power in pounds sterling of Table XXIII. Wages of Trainmen; General Expense. Each of these items would be nearly the same for either of the alternative routes, and can be safely omitted from the comparison, although the item of train-wages, on the basis of a train-mile distribution, would affect the B lines somewhat adversely, because of their greater length, as compared with the A lines. Car Repairs and Renewals. This has been obtained from the returns of the New Zealand railways, with a percentage of increase due to rise and fall and curvature, in the same manner as in my first supplementary report, correction being made to agree with recent data, now at hand. With the rate per trainmile from the last column of Table XVIII., and the total number of train-miles per annum, for each case, the total yearly cost of renewals and repairs of cars is found as given in Table XIX. If the trains brought up to Bealey are rated for and hauled by an 85-ton locomotive, 340, 477, and 681 trains per annum would handle the same business as 500, 700, and 1,000 trains per annum with Class B locomotive. The total number of car-miles and of ton-miles would therefore remain the same, while the train and locomotive miles per annum would be less, and the weight of the trains and locomotives would be proportionally greater. It is therefore assumed that the main-tenance-costs would be essentially the same for the same volume of traffic with either locomotive. There are no data at hand from which the exact effect on maintenance-charges of handling the same tonnage in heavier and fewer trains can be determined, but it is evident that there are factors which would balance each other —that is, some items would be larger for the heavier trains and others would be larger for the lighter trains. The net result cannot vary enough to materially affect this comparison. Comparison of Working-costs. Adding together the annual cost of motive power from Table XXIII., the cost of maintenance of way from Table XVII., the cost of renewals and repairs of cars from Table XIX., and the cost of tunnel-ventilation from Table XXV., we have the total of all the variable elements in the cost of working these alternate lines as shown in Table XXVI. It will be seen from this table that the cost of working the several lines considered does not differ greatly. The extreme difference between Line 83, the highest, and Line A 4, the lowest, is only £14865 sterling per annum for 500 trains each way per year. For 1,000 trains each way per year the difference is £2986 sterling. Adding interest at 3J per cent, per annum on the cost of construction given in Table 1., we have the sum of all the variable elements in the total annual charges in pounds sterling. These results are shown in Table XXVII. They distinctly indicate that Line A 4 is the best. Line B 1, which is the one mentioned in my first report as likely to be the best of those considered therein, follows line A 4, ranking as a close second in the order of merit. If the railway was to be built purely as a commercial enterprise it would doubtless pay to give more favourable consideration to line B 1, because, on account of its shorter summit tunnel, the railway could be completed and begin to earn money at an earlier date than in the case of either of the other routes. But it would not be good policy, nor pay the Government in the long-run, to build any but the best line, especially as such line is estimated to cost the least. Line A4is an ideal route, considering all the points of the physical situation. It keeps clear of the hill-sides and of all possibility of interruption by slides, or danger from falling rocks, is short, and is almost, although not wholly, free from curvature. In fact, a better route is not feasible, and I am glad to recommend it to you. No attempt is made herein to estimate the extra cost of maintenance of way of either route, due to slides or wash from slopes, &c, this item being a matter of pure conjecture. It is not at all improbable, however, that this would, during some years, so increase the aggregate working-cost of line B 1 that it would exceed that of line A 4 in the sum of £1,000 to £1,200 sterling, as estimated by the superintending engineer. I take pleasure in saying that the Government engineers have shown ability in their work in these premises, and in developing the data sent me. Temporary Line. In my reports I have mentioned the Shay locomotive as a practicable machine for the operation of steep gradients. It is slow in speed, but very positive and safe in its action. It can, however, easily make eight miles per hour, and does do so every day in the year on the Mount Tamalpais Scenic Eailway in California. The following data relating to this railway will be of interest to you. The Railway. The railway is of standard gauge, and was built in 1896 under the engineering direction of Mr, George M. Dodge, of San Eafael, California. The line is a trifle less than eight miles and a quarter in length, with a total curvature of about 15,000 degrees, and a total ascent of about 2,500 ft.

64