GREAT BRITAIN, II gaa 255 her gross barter terms of trade, but has nothing to do with the net barter terms. The same is true, mutatis mutandis, of fresh loans and investments. These also affect the gross barter terms, but not the net barter terms. Of course they affect the gross terms the other way. The export of capital means that goods are sent out, but are not expected or stipulated to be paid for in the current year. For the time being the goods go out from Great Britain, but neither merchandise nor money comes in as payment. Eventually, of course, income will be paid ; and it will come in the form of imported goods. Some annual payment may begin the very next year, and presumably a return flow will set in (unless the investment happens to be a complete failure) within a very few years. Inthe end, even the principal may be repaid; and then the payment made to Great Britain will be heavy and her merchandise imports will be very greatly swelled. But all these transactions, with their trend first one way and then the other, have a direct effect, year by year, on the gross terms of trade only. They are thus different from such transactions as are involved in shipping services; these being part of the immediate volume of British products which goes out from year to year and is paid each year in the form of inflowing imports. It follows further that in any strict calculations of the net barter terms, and of changes in those terms, regard should be had not only to the volume of shipping services, but also to their prices — to the rates of freight. The index of export prices which has been used in constructing the chart and the tables does not include shipping, nor such minor items as banking and insurance service. Therein it obviously fails to be inclusive. Doubtless the materials exist from which shipping rates could be computed, and the index of export prices thus made more accurate. I have not endeavored to modify in this way the price indices, but have taken them as they appear in the British sources — figures, that is, derived from the prices of merchandise alone. Tt is possible that corrected indices would show a different trend, but not probable. I feel reasonably confident that no differences would appear such as to invalidate the general conclusions here drawn. The point