with the amount the foreigner spends in America, as indicating the ratio of foreign to domestic trade for the purpose in hand. We subtract, in other words, from the figure for total income the figure for _imports_. Then we compare the residue with the figure for _exports_, and get our ratio of foreign to domestic trade. The export and import figures must first, however, be reduced to a _retail_ basis. That is, assuming that wholesale prices are two-thirds of retail prices, we add 50% to the figures for exports and imports (which are wholesale figures) before making the subtraction and the comparison. The ultimate consumer, both in Europe and America, pays for imports and exports on a _retail_ basis.[304] This method, applied to the figures for 1910, gives us a ratio of about 10:1 for domestic to foreign trade--the lowest percentage for foreign trade which we shall find for any year in the period investigated, 1890-1916. This comparison is still unfavorable to foreign trade. Domestic trade, in our figures, includes savings and investments, including investments made by Americans abroad. Import figures are marred by undervaluations, exports are not all counted, and the figures for exports and imports do not include foreign investments in America. American investments abroad should not be counted as part of domestic trade. Moreover, our figures take no account of travellers' expenditures, or of services performed by professional men of one country for men in another, or of certain other "invisible items." But while this makes our percentage for foreign trade too low for all years, it probably does not greatly upset the results for yearly variations in the ratio except for the year 1916, when the figure for domestic trade is left decidedly too high, and the ratio for foreign trade is too low, as compared with previous years.