574 MINERALS YEARBOOK
 The most significant feature of the chart is that cement prices, both composite
and Buffington, after dropping precipitously dur ng the first few months
of 1931, began to recover during the latter half of 1932. By the early part
of 1933 the composite price curve had reached a level, comparable to its
average position with respect to building materials, and the Buffington curve
was almost as high.
 Regardless of the upward turn in* prices in midsummer, however, the average
factory value per barrel of cement dropped from $1.11 in 1931 to $1.00 in
1932, a decline of 9.9 percent. The monthly average of the composite price
index in 1932 decreased only 3 percent from 1931. From 1928 to 1932 the average
factory value per barrel of cement declined 36 percent, whereas the composite
price index dropped only 19 percent. This discrepancy apparently reflects
the difficulty of compiling a price index which reveals prices of actual
sales rather than quotations. It is generally known that in 1932 sales were
made in the Middle West at prices appreciably below prevailing quotations.
This problem is unusually real during periods of keen competition such as
the cement industry has experienced in recent years.

SPECIAL CEMENTS

 Natural cement competed sharply with portland cement during the early years
of its manufacture, but since 1900 portland cement has dominate.d the industry.
Ten years ago portland cement comprised about 99 percent of total shipments;
the other 1 percent consisted of natural, masonry, and puzzolan cements,
used primarily for mortar and stucco. Shipments of natural, masonry, and
puzzolan cements amounted to 1,226,850 barrels in 1931 but apparently declined
about 55 percent in 1932. This is considerably more than the comparable decrease
for portland cement, but it must be remembered that in 1932 the principal
market for portland cement—concrete paving—dropped only 28.0
percent from 1931 whereas the volume of building construction declined 57.5
percent.
 During the past 10 years numerous cements selling at premium prices and
recommended for special purposes have been developed. Since 1927 the Bureau
of Mines has attempted to collect statistics to show' their growth, but the
problem is difficult because of the confusion which stifi exists regarding
the exact classification of these so-called "special cements." However, a
logical break-down was adopted for 1931 statistics, and this classification
will be continued even if differences are noted in comparing figures with
those for preceding years.
 Production of high-early-strength portland cement in 1931 amounted to 1,361~,468
barrels, whereas shipments were 1,422,633 barrels valued at $2,278,236. This
quantity represents returns from 16 plants, some of which reported production
of high-early-strength cement for the first time in 1931, indicating that
interest in the material is increasing. The average factory value of the
product was $1.60 a barrel, $0.49 higher than that of standard portland cement.
Obviously, only cement selling at premium prices has been included in the
total, but information from producing companies is insufficient to determine
accurately whether or not all the cement thus reported actually conformed
with the temporary specification of A.S.T.M., 0 74—30 T. The adoption
of a permanent specification for high-