employed in the Aldrich report consisted in giving to the various commodities or groups of commodities an importance based upon their consumption in normal families. There are, of course, many commodities whose importance can not be measured by this method. It has been thought best in the present series of index numbers, after a careful consideration of all methods of weighting, simply to use a large number of representative staple articles, selecting them in such a manner as to make them, to a large extent, weight themselves. Upon a casual examination it may seem that by this method a comparatively unimportant commodity-such, for instance, as cotton-seed meal-has been given the same weight or importance as one of the more important commodities, such as wheat. A closer examination, however, shows that cotton-seed meal enters into no other commodity under consideration, while wheat is not only quoted as the raw material, but enters into the two descriptions of wheat flour, the two descriptions of crackers, and the three descriptions of loaf bread.

Material changes in the description of three articles were made in 1902, of two articles in 1903, and of one article in 1904.

For two of these articles the trade journals no longer supply regular quotations, the manufacture of the particular grades of three previously quoted has been discontinued by the establishments heretofore furnishing quotations, and for one article the substituted description more nearly represents the present demands of the trade.

The articles in which changes occur are "Hosiery: men's cotton half hose, seamless, fast black, 20 to 22 ounce, two-thread goods," for which has been substituted "Hosiery: men's cotton half hose, seamless, fast black, 20 to 22 ounce, single-thread goods;" "Leather: harness, oak, country middles, 14 pounds and up (except overweights, 20 pounds and up)," for which has been substituted "Leather: harness, oak, packers' hides, heavy No. 1;" "Sheetings: brown, 4-4, Stark, A. A.," for which has been substituted "Sheetings: brown, 4-4, Massachusetts Mills, Flying Horse brand;" "Trouserings: fancy worsted, 22 to 23 ounce, worsted filling, wool and worsted backs," for which has been substituted "Trouserings: fancy worsted, 21 to 22 ounce, all worsted warp and filling, wool and worsted back;" "Underwear: shirts and drawers, white, merino, full-fashioned, 52 per cent wool, 48 per cent cotton, 24 gauge," for which has been substituted "Underwear: shirts and drawers, white, merino, full-fashioned, 60 per cent wool, 40 per cent cotton, 24 gauge;" and "Shingles: white pine, 18 inches long, XXXX," for which has been substituted "Shingles: Michigan white pine, 16 inches long, XXXX."

In making these substitutions articles were supplied corresponding as closely as possible to those which had been previously used.

Some explanation of the method of computing the relative price of these articles is necessary, and shingles will be used as an illustration.

It must be understood that during the years when 18-inch shingles were quoted they were assumed to represent the several lengths and grades of white pine shingles; that is, that the course of prices of a standard grade of white pine shingles in an index number of prices fairly represents the course of prices of the various grades of white pine shingles. Therefore when it became necessary to substitute, in 1902, the 16-inch shingles for the 18-inch, prices were secured for 16-inch shingles for both 1901 and 1902, and it was found that the average price for the year 1902 was 9.96 per cent above the average price for the year 1901. The relative price of 18-inch shingles in 1901, as shown in Table IV, was 111.9 (average price for the 10 years, 1890 to 1899, equals 100), and if 18-inch shingles represented white pine shingles at that time and 16-inch shingles now represent the class, shingles (shown by the increase in price of 16-inch shingles) advanced, in 1902, 9.96 per cent above the price in 1901, and the relative price in 1902 was therefore 109.96 per cent of 111.9, the relative price in 1901, which gives 123 as the relative price in 1902. The same method was followed in computing relative prices for each of the months of 1902. The average price in 1904 was 2.05 per cent below the price in 1903; therefore the relative price in 1904 was 97.95 per cent of 125.1, the relative price in 1903, which gives 122.5 as the relative price in 1904. The same method was followed for hosiery, leather, and sheetings. For trouserings and underwear the exact grade quoted for 1903 was not manufactured in 1902. The manufacturer of trouserings, however, estimates that one-half the advance in price over the price for the grade quoted for previous years was due to the fact that it was a better article and the other half to the advance in price of material and cost of manufacture. The advance was $0.1125 per yard over the price in 1902; one-half of this, $0.05625, was added to the 1902 price of the 22 to 23 cunce trouserings to secure a theoretical 1902 price for the 21 to 22 ounce trouserings, and the 1903 relative price was then computed as above. Underwear was arbitrarily given the same relative price in 1903 as in 1902, as the all-wool underwear manufactured by the same firm showed no change in price. The 1904 relative prices of trouserings and underwear were found in the same way as was explained above for shingles.

Table III.-Relative prices of commodities in 1904, pages 507 to 518.— This table is taken from Table II and shows the relative prices of each of the commodities included there. In this table similar commodities have been grouped, and the average of the relative prices shown for the commodities in each subgroup and in each of the nine general groups. The averages in all cases were found by dividing the sum of the relative prices by the number of commodities in the group under consideration. It should be borne constantly in mind that the term commodity is used here and elsewhere in a specific sense, "native"

and "Western" sheep, for example, being considered different commodities. The method of securing average relative prices in this and other tables was as follows: The average relative price of cattle was found by adding the relative prices of the two grades of cattle and dividing the sum by 2. The average for hogs was found in the same manner, and also the average for sheep. The average for live stock was found by dividing the sum of the relative prices of the two grades of cattle, the two grades of hogs, and the two grades of sheep by 6, the total number of different deseriptions of commodities or series of quotations in the live-stock group. The average relative price of each of the nine general groups was found by dividing the sum of the relative prices of the different descriptions of commodities for each month by the number of these commodities or series of quotations considered. The sum of the relative prices in January, 1904, of the commodities shown under the general group, "food, etc.," for example, is 5,635.8, which amount divided by 53, the number of different descriptions of commodities or series of quotations considered in that group, gives 106.3 the average for the group "food, etc.," for January, 1904. As explained in the discussion of Table II, it was impossible to secure quotations during all of the months of the year for 5 of the 259 articles. In order of arrangement these are herring, buckwheat flour, onions, potatoes, and sheetings-bleached, 10-4, Atlantic. In presenting monthly relative prices for these articles a nominal relative price (which is the same as the relative price for the month in which the article was last quoted) has been entered in this table for the months for which no price quotation is shown in Table I. This nominal price enters into the average for the subgroup, the general group, and "all commodities" for that month.

The average relative price of each of the nine general groups for each month of 1904 and for the year 1904, and the average relative price of all commodities for each month and for the year, are shown in the following summary:

SUMMARY OF RELATIVE PRICES OF COMMODITIES IN 1904, BY GROUPS.
[Average price for 1890-1899=100.]

In this table the average relative prices of farm products are based on 16 articles; of food, etc., on 53 articles; of cloths and clothing, on 76 articles; of fuel and lighting, on 13 articles; of metals and implements, on 38 articles; of lumber and building materials, on 27 articles; of drugs and chemicals, on 9 articles; of house furnishing goods, on 14 articles; and of miscellaneous, on 13 articles. The average relative prices of all commodities are based on 259 articles.

The table shows that the group of farm products reached the lowest average in January and the highest in March; that of food, etc., the lowest in June and the highest in December; that of cloths and clothing, the lowest in November and the highest in February; that of fuel and lighting, the lowest in July and the highest in January; that of metals and implements, the lowest in September and the highest in December; that of lumber and building materials, the lowest in November and the highest in June; that of drugs and chemicals, the lowest in November and the highest in May; that of house furnishing goods, the lowest in February, March, and April, and the highest in January; while in the miscellaneous group the lowest average was reached in November and the highest in March and July. All commodities combined reached the lowest average of the year in October and the highest in March.

The course of prices during the months of 1904 as represented by the 259 commodities is clearly shown in the graphic table which follows:

RELATIVE PRICES OF ALL COMMODITIES IN 1904.

[Average price for 1890 to 1899-100.]

RELATIVE JAN. FEB. MAR. APR. MAY. JUN. JUL. AUG. SEP. OCT. NOV. DEC.

PRICES.

116

114

112

110

108

106

104

102

Many students of price statistics desire to distinguish between raw commodities and manufactured commodities, or those which have been prepared for consumption by the application of manufacturing processes and in which manufacturing labor forms a considerable part of the cost. To meet the wishes of this class of readers, therefore, the commodities included in this price series have been divided into the two classes, raw and manufactured, and simple averages made for each class. Of course, hard and fast definitions of these classes can not be made, but the commodities here designated as raw may be said to be such as are marketed in their natural state and such as have been subjected to only a preliminary manufacturing process, thus converting them into a marketable condition, but not to a suitable form for final consumption, while the commodities here designated as manufactured are such as have been subjected to more than a preliminary factory manipulation and in which the manufacturing labor cost constitutes an important element in the price. In the group designated as raw are included all farm products, beans, coffee, eggs, milk, rice, nutmegs, pepper, tea, vegetables, raw silk, wool, coal, crude petroleum, copper ingots, pig lead, pig iron, bar silver, spelter, pig tin, brimstone, jute, and rubber-a total of 50 articles.

All the other articles are classed as manufactured commodities.
The table follows:

RELATIVE PRICES OF RAW COMMODITIES, MANUFACTURED COMMODITIES, AND ALL COMMODITIES, IN 1904.

The raw commodities reached the lowest average of the year in October and the highest in February; manufactured commodities reached the lowest in September and the highest in March. The average for raw commodities ranged from 17.3 per cent to 23.6 per cent above the base price, while the average for manufactured commodities ranged from 10.3 per cent to 12.5 per cent above the base price.