classes of traffic could be distributed with accuracy, what was the practical good of it? It might be proved, for instance, that second class passengers were carried at a loss of so many thousands a-year by attributing to them their full share of the cost of every kind that they incurred. But a railway man would argue that his second class passengers paid him 100,000l. a-year more than they would pay if they travelled third class, so that if it only cost him 80,cool. more to carry them in second class carriages than in third, he made 20,000l. a-year by the traffic. Therefore even if by averaging the expenditure over all three classes it was proved that the second class passengers yielded a minus quantity of profit, that still did not prevent the railway man having an interest in carrying them, still less did it prevent its being in the public interest that those who wished to go second class and liked to pay for it should be allowed to do so. He confessed therefore that not only was he convinced it was impossible to allot the working expenses between goods an passengers, and still more to divide them further between different classes of passengers, different classes of goods, and cost of terminals and conveyance, but that, even if it were possible, he believed the division would be of no practical utility.

Mr. C. J. OWENS, referring to Mr. Price-Williams' statement that for a certain period the London and North Western gave certain statistics, but then discontinued them, said that this company was progressive and not retrograde; and had they believed that the figures they were giving were trustworthy and valuable, no doubt they would have continued them. But, speaking with a lifelong experience of railway work, he said most emphatically that the division of the expenses of the working of a railway as between goods and passenger traffic, and still further as between various classes of the two main descriptions of traffic, was an impossibility. The same staff at all the small stations on a railway such as that he represented, which were 60 or 70 per cent. of the whole, did all the work there was to be done, and how could the expenses of the staff at such stations be divided between the goods clement and the passenger element? Then, again, how were the important expenses of signalling to be distributed? Certainly not in the ratio of the receipts, because the passenger traffic had to be specially protected in accordance with the regulations of the Board of Trade. Further on in the paper the author mentioned losses incurred in the working of the passenger traffic. In this matter he fully agreed with Mr. Acworth. They were told on p. 483 that there was an actual loss of 28. per first class passenger, or 1d. per train-mile. Such a statement would assume that if they did not carry these first class passengers they would effect a great saving. But how could they allot to this particular class of traffic the expenses of signalling, running the trains, locomotives, stations, and staff? They were all permanent expenses, and would be just the same whether the whole traffic was first class or divided into three classes. If the expense of providing improved (i.e., first or second class) travelling accommodation was more than counterbalanced by the increased rates charged, then those classes of

traffic would be worked at a profit. On p. 499 was given the average station terminal, which was arrived at by dividing the terminals of all the various classes and dividing them by the number of classes. This would be a perfectly accurate assumption if they carried 1,000 tons in each of the various classes. But, dealing with goods traffic, i.e., Class C and classes 1 to 5, 60 per cent. of the total tonnage would be in Class C, which gave the lowest terminal, and somewhat less than 1 per cent. would be in class 5, which gave the highest terminal. By the method adopted therefore a figure was obtained which was wide of the actual fact. He agreed with Mr. Price-Williams when he referred to the inadequate allowances in some cases made for terminal services. When this matter came before the Board of Trade the question was discussed as to what was to be allowed for loading furniture. That came in class 5, the figure being 18. 8d. per ton, which was accepted by the railway companies, although he was convinced that none of the large furnishing houses could perform the service for that amount. The terminal services and conveyance rate were treated separately; but what railway companies aimed at in the division of the rates was that in the aggregate they should get what was fair. In some instances they did not get what was fair in terminals, but they got compensation in the conveyance rate, and vice versa. Their main endeavour had been to maintain the status quo.

The CHAIRMAN (Sir COURTENAY BOYLE, K.C.B.) wished to safeguard himself by saying that if he did not refer in detail to the majority of the points mentioned in the paper, he must not be held to acquiesce entirely in the accuracy of the arguments set forth with so much care and labour. The question of railway rates was very full of difficulties and intricacies, with which a prolonged study could alone enable an expert to grapple. Just as an instance he might mention that in one place the author said that minerals came under the category of A, or the lowest class of merchandise traffic, but he thought the last speaker would bear him out that a good many minerals came under Classes B and C in the classification. Then ou p. 498 Mr. Price-Williams calculated the mean of the rates for distances within the limits given, and worked out the mean for 20 miles at 19d., for 50 miles at 42°5d., and for 100 miles at 50d. But he ventured to think that Mr. Price-Williams did not quite appreciate the effect of the schedule to the Confirmation Act of 1891. The charge for the first 20 miles ran through the whole of the conveyance, and consequently it was only for 30 miles not for 50 that the rate was 0.85d., and therefore that 42'5d. in the third column should be 44 5d., and the 50d. ought to be 69.5d. One of the most important suggestions was that the railway companies would do statisticians a most important service if they could tell them what was the cost of a ton-mile, and this point had been urged by the Board of Trade as well as by traders. Undoubted difficulties had been urged, but he thought that if it was understood that the figure given by the railway companies for a ton-mile was an estimate and not an arithmetical fact on which

argument might be based, some more information might be given than was given at present. He would suggest to the able representatives of the railways that they should once more consider whether it was not possible to give an estimate of what the cost of a ton-mile was. The author of the paper doubted whether any adequate allowance had been made in the rates and charges authorised in respect of goods on mineral traffic for terminal expenses. They were told repeatedly that the terminal allowances were extremely high and bore hardly on the trader. They were also told by the railway companies that the full maximum powers of charging terminals were very seldom exercised. Lastly, he would say that railway statistics, like other statistics, must be used with a knowledge of what they really were. They must be treated as more or less estimates, useful as a guide but not indicating a perfectly accurate fact. He concluded by moving a cordial vote of thanks to the writer of the paper for the extreme care and patience he had taken in its compilation.

The following reply to the various criticisms made on the paper during the discussion was subsequently forwarded to the Editor by the author.

Mr. Price-Williams wished to express his appreciation of the value of the criticisms upon the paper. As regards Mr. Acworth's observation, that if the author had discovered a method of getting the requisite information for the apportionment of the expenses between the passenger and goods traffic, what need was there to trouble the railway companies for further information, it was only necessary for him to point out that it was the absence of this fuller and more detailed information in the railway companies' reports, which obliged his having recourse to the method he had adopted. That the necessary information could be furnished by the railway companies was admitted and exemplified in the case of the London and North Western Railway reports during a long period of years, and in a still more complete manner in the Government annual reports of the Belgian railways, the reports of some of the American railway companies (notably the New York Central and Hudson River), and also in the admirable "administration reports" of the Indian railways, for which, in the latter case, they were largely indebted to a distinguished Fellow of this Society, Sir Juland Danvers. In all the railway reports alluded to, the ton-mile and the passenger-mile units were important features, and a great step would be gained were the railway companies in this country to furnish in their reports the requisite data to enable these important factors to be determined.

It must be admitted that the results arrived at, even with the aid of the fullest information, would necessarily be, as Sir Courtenay Boyle had pointed out, only approximately correct; sufficiently so, however, to admit of reliable conclusions being arrived at as to the relative cost of working on different railways,

due regard being had to the nature of the traffic and the character of the respective railways, and still more reliable when applied for the purpose of determining the relative working expenses attributable to each of the three great branches of traffic on a particular railway system like that of the London and North Western, where any slight errors in the apportionment such as those alluded to in the discussion would, as was admitted by the late Sir George Findlay, be entirely neutralised.

No one acquainted with the working of American railways and the admitted absence of anything like uniformity in the mode of dealing with their accounts, would be at all surprised to learn from Mr. Acworth that the "Inter-State Commerce Commission " which was instituted in 1887 for the development of railway statistics under Federal control and for the establishment of a uniform system of accounts, had decided to discontinue their attempts to distinguish the expenses attributable to passenger from those attributable to goods traffic.

That this decision was the inevitable result of their inability to get the American railway companies to adopt a uniform system of accounts there could be little doubt, the Commissioners however had evidently not abandoned the hope of doing so, as in their report they expressed their opinion "that uniformity of accounting would never be obtained in this manner, and that the Commission would ultimately be obliged to exercise the power which Congress had reposed in it to prescribe for the carriers (i.e., the railway companies) a uniform system of accounts, and the manner in which such accounts were to be kept."

The few following extracts from the Commissioners' report for 1892, to which allusion had been made by Mr. Acworth, afforded some idea of the great and exceptional difficulties encountered in their efforts to bring this about, and sufficiently explained the reasons which at present rendered it so difficult to institute any reliable comparisons as to the relative cost of working the various railway systems in the United States, still more so to attempt accurately to apportion the working expenses as between the passenger and goods traffic.

*

"So far as the general principles of accounting were concerned there appeared to be substantial harmony, but there existed a sufficient diversity in matters of detail to destroy any comparisons which might be made on the basis of the statistics collected by the several States. The Convention, recognising the necessity of uniformity, passed a resolution to the effect that it is the sense of this Convention that a uniform method of collecting and publishing statistics, both as to time and matter, should be adopted.”

**

"The chief occasion for uncertainty at the present time is found in the fact that the carriers (the railway companies) do not themselves follow uniform methods of accounting."

"It is a noteworthy fact that the percentages of items of cost of sub-classes of operating expenses are not, in many cases, the

same when compiled from the official reports and from the reports from railways to stockholders."

*

*

*

"It is no secret that under present conditions it is exceedingly difficult for the shipper (freighter) whose rights are disregarded by a railway corporation to secure quick and speedy relief; and on this occasion shippers conceive their interests to depend upon the good will of railway managers rather than upon the protection of commissions or courts, and consequently are reluctant to bring cases for the defence of their established rights."

"In order that the law against discrimination in rates may become effective there must be created a uniformly organized and uniformly administered railway system. Managers cannot be allowed the liberty of adopting unusual methods of business, nor lawyers the right of urging before commissions peculiar policies of management as a defence for unusual methods."

"Now the first step towards the establishment of uniformity of management is to establish uniformity in accounts, and to take from railway officials the right of adjusting their accounts in an arbitrary manner. The railway laws of this country are not at present capable of easy execution, because of the difficulty of securing evidence against discrimination'-a difficulty which in large measure is due to the numberless and complex methods by which railways do their business."

"The question involved in this controversy is not simply commercial in character; it is at the same time a question of public policy, and as such, like all questions of a political character, demands the fullest and completest knowledge respecting it. If the theory that railway rates should bear some relation to cost of service be accepted, it is essential that the facts pertaining to cost of service should be known.""1

"It will probably be said that a bureau of statistics and accounts outlined in the above recommendations contemplates an organization on a basis so broad as to preclude successful administration. The reply to this objection is that it is not magnitude of work assigned to a bureau which makes it difficult for administration, but complexity in the elements with which the bureau is called upon to deal; and provided only there be established uniformity in the general accounting of corporations engaged in inter-state traffic, and provided that Congress is willing to grant ample means and authority, there is no reason in the nature of the case why the plan outlined in this report may not be realized."

The main difficulty experienced by the Inter-State Commerce Commission had in this country been already surmounted, as all the railway companies' reports, and the returns they were required

11 The italics are my own.-P.-W.

VOL. LIX. PART III.

2 M