Investigation upon this latter head brought out three important facts. The first was that one great source of expense was to be found in what is technically called “taxing” the letters, that is, ascertaining and marking the postage to be charged on each; the second, that great expense likewise arose from complicated accounts, postmasters having to be debited with unpaid postage on letters transmitted to their offices, and credited with their payments made in return; while they again had to receive and check the payments of the letter carriers, who, it must be remembered, received, at that time, from the public, almost all the postage paid; the third, that the cost of delivering letters, great as it inevitably was, was much augmented—indeed, save in rural districts, more than doubled—by being saddled with the collection of postage. It further appeared that these expenses must increase in something like direct proportion to increase in the number of letters. These conclusions led me to perceive that for the effectual reduction of expense it was necessary to obtain simplicity of operation, and therefore to reduce the prodigious variety of rates (then extending, on single inland letters alone, to upwards of forty), and further, to adopt means to induce prepayment, so as to save the time at once of the letter carriers, of the clerks with whom they had to account for postage received, of the provincial postmasters, and, lastly, of the clerks at the central office. In considering how far the variety of rates might be reduced, I was naturally led to inquire what proportion of postal expense proceeded from the conveyance of letters between town and town, and further, how far such expense, whatever it might be, varied in relation to distance. On pursuing this inquiry, I arrived at results so startling that nothing but the most careful verification could satisfy me of their accuracy. I first perceived that the expense of such conveyance, which one would naturally suppose to be very great, was in fact, when divided by the number of missives, very small.