bers of several industries average about 5 per cent of the total annual value of production. Some individual concerns have experienced a greater percentage saving.

Carrying charges for handling and stocking items vary with different industries from 2 to 15 per cent. This fact considered in conjunction with the high average percentage of variety reduction effected and other advantages inherent in simplified practice leads to the assumption that 5 per cent of the total value of production is a conservative basis for its evaluation.

Using this method of calculation, it was estimated that 40 simplified practice recommendations, or 41 per cent of the total 98 now in effect, were saving their respective industries over $134,600,000 for the calendar year 1928, and cumulative yearly savings from 1922 to 1928, inclusive, for these 40 recommendations amounted to over $595,300,000.

SPECIFICATIONS

Encyclopedia of Specifications. Standards and Specifications for Nonmetallic Minerals and Their Products, being the second volume of the encyclopedia series, is now in manuscript. The first volume of this series, published in 1927, is entitled "Standards and Specifications in the Wood-Using Industries."

A survey is being made to determine scope of the third volume, which will probably deal with standards and specifications for metals and their products (except machinery and vehicles).

Directory of Specifications.-The National Directory of Commodity Specifications, issued in 1925, is being revised.

Publications. Safeguarding Manufacturing Economies, A. S. McAllister, United States Daily, October 8, 1928:

Applying Government Specifications to Utility Purchasing, A. S. McAllister, Public Utility Purchasing, 1, No. 1, February, 1929.

Performance Specifications-An Effective Compromise between Dimensional Standards of the Seller and Quality Specifications for the Buyer, A. S. McAllister, Purchasor, Vol. VIII, No. 2, February, 1929.

SPECTROSCOPY

Searching for a Better Standard Wave Length.-In 1927 the International Conference of Weights and Measures provisionally adopted the red radiation from cadmium (6438.4696 A) as the primary standard of wave length, but recommended that the krypton line (565 mμ) be investigated with a view of its eventual substitution for the red ray of cadmium.

In accordance with this recommendation, the first spectrum of krypton was investigated; its description was revised, and its structure was determined. As certain properties of spectral lines (pressure effect, reversibility, Zeeman effect, relative intensity) are known to be related to the spectral terms (atomic energy levels), it appeared desirable first to classify the spectrum lines. Then, theoretical considerations can be used in discussing the suitability of a line for use as a standard wave length. An analysis of the krypton spectrum shows that theoretical objections can be raised against the proposed yellow-green line (565 mμ). This line arises from the term combination 183-3p10 (3P-3D1), a transition between levels with small inner quantum numbers 0 and 1. This accounts for its relatively low intensity; the intensity is considered insufficient for general practical

use.

Furthermore, the final state 18, is a metastable level and lines involving such metastable levels are the first to show absorption or reversal. The yellow line of krypton (5,870.92 A) is estimated to be at least twenty times as intense as the proposed yellow-green line; it arises from the term combination, 184-2p2 (P1-'D2), thus involving stable levels with larger inner quantum numbers, 1 and 2.

A comparison of the energy levels of the krypton and the cadmium spectra shows that the krypton lines which have been suggested as standards occupy roughly the same relative positions in the krypton diagram as the provisional standard does in the cadmium diagram. Excepting the lines connected with the long-lived metastable states of krypton, the remaining combinations may be expected to be no more sensitive to reversal and pressure than the cadmium standard itself. Other things being equal, it appears that the yellow line is the best kyrpton standard, but it will be necessary to investigate its homogeneity more carefully before it can be recommended as a primary standard.

Publication.-The First Spectrum of Krypton, W. F. Meggers, T. L. de Bruin, C. J. Humphreys, B. S. Jour. of Research, 3, No. 1, p. 129, July, 1929.

STONE

Slate Investigation. In connection with a general study of the physical properties of slate, about 60 samples have been collected from quarries in Maine, Vermont, New York, Pennsylvania, Maryland, Georgia, Tennessee, and Arkansas. Approximately 2,800 tests have been completed.

The results to date indicate the following ranges and average values:

Abrasive Hardness.-The abrasion testing apparatus has been further improved and data on marble, limestone, sandstone, granite, and slate have been obtained as follows:

ortonal to the reciprocals of the volumes une the abrading being done with No. 60 artiter conditions.

meiding stones that have heretofore 84127 permanent have been found to reach TWIN 60 years, a special study has been he rate of externally applied stone preservander various conditions. A study of the As applications was previously inaugurated

SUGAR

merous inquiries, a table showing the weights set and weights per cubic foot of sugar (sucrose) is een prepared. This table comprises weights per gallon in air and in vacuo, weights oot in air, specific gravities 20°/20° and Saute, all at intervals of 1° from 0° to 95° Brix. ased on the density values of Plato. The dexed from the table of Bates and Bearce.

Weights her U. S. Gallon and Weights per Cubic WCC.F. Snyder and L. D. Hammond, B. S. Circ. 22.714; July 27, 1929; and The Planter and 581; June 29, 1929.

PESTS

ced Dung Fiscal Year-An itemized statement to bureau during the fiscal year ended June 30, ne number of determinations and the fee value,

Numbers of test items, determinations, and fee value for tests completed during the fiscal year ended June 30, 1929-Continued

Includes fee value of $23,821 for sampling and shipping 1,044,093 barrels of cement.

3 Of these totals 229, 249 determinations were for the public, fee value, $75,153.75; 348,510 determinations were for the Government department and State institutions, fee value, $427,628.38; 22,761 determinations were for the bureau, fee value, $41,620.20. The number of test items and determinations necessary in connection with the bureau's own work of research and standardization, with the resulting fee values, is not included in these totals.

TEXTILES, DYES, CLEANING, ETC.

Standardization of Knit Underwear.-The Associated Knit Underwear Manufacturers of America maintain a research associate in the bureau. Standard measurements for the following types of underwear were developed during the year and were adopted by the association:

1. Men's rayon knit athletic shirts (pull-over).

2. Men's ribbed (1 by 1) cotton union suits (no sleeves, three-quarters length, button front).

3. Men's ribbed (1 by 1) cotton union suits (no sleeves, three-quarters length, button shoulder).

4. Men's rayon ribbed knit athletic union suits.

5. Men's rayon flat knit athletic union suits.

6. Boys' cotton ribbed knit athletic union suits.

7. Boys' cotton flat knit athletic union suits.

8. Boys' knit athletic shirts (pull-over).

9. Women's ribbed (1 by 1) worsted, worsted-merino, wool, and wool-cotton union suits (high neck, long sleeve, and ankle length).

10. Women's ribbed (1 by 1) cotton union suits (Dutch neck, elbow sleeve, ankle length, button front).

11. Women's ribbed (1 by 1) cotton union suits (low neck, no sleeves, knee length, built-up shoulder, tight knee).

12. Women's ribbed (1 by 1) cotton shaped vests (low neck, no sleeves, builtup shoulder).

13. Women's ribbed (1 by 1) cotton drawers (knee length, tight knee).

14. Standard practice for spacing shoulder straps on women's bodice-top union suits and vests.

Wool Meter. A simple portable instrument for determining the average diameter of wool fibers has been devised. Data quickly obtained with it differ by less than two microns from those obtained from laborious, time-consuming, microscopical measurements.

Determination of Strengths of Yarns.-A preliminary survey of projects regarding rayon properties showed that the methods used for determining strength and stretch of yarns were inadequate. A study of the subject later enlarged to include yarns of fibers other than rayon has resulted in the development of a method that has all the advantages of the lea method, and a higher degree of accuracy. It is known as the equitension lea method, since the yarn is wound under uniform tension on a specimen holder, which is then inserted between the jaws of the testing machine and the specimen broken. Very consistent results have been obtained both in stretch and strength of yarns of rayon, cotton, wool, etc. A suitable apparatus has been designed and the procedure formulated. This work is practically completed and the information obtained will be published.

Investigation of Dyes. Quantitative measurements of the spectral absorption of solutions of the methyl and of the dimethyl derivatives of benzeneazophenol have been made and the relation between chemical constitution and absorption studied. Twelve oil-soluble dyes of purity satisfactory for standard spectral absorption measurements have been prepared.

A comprehensive report on the fading of dyed textiles in daylight and in carbon arc light has been published, and a report on the fastness of lithographic ink pigments to light prepared. Work has been continued on the selection of standards of fastness of dyed textiles, both to light and to washing.

Forecasting of Resulting Color.-In the final analysis, the value of a dye depends on the resulting color, or spectral reflection, of the materials to which it has been applied. Spectrophotometric measurements of reflectivity yield quantitative data on the color and can be used to advantage in testing dyes, in the study and control of dyeing operations, and in determining the fastness of dyed materials. It has been found that the reflectivity, and thus indirectly the color, of any dyeing made with a practicable amount of the dye can be predicted from the measured reflectivities of a very few dyeings, and that the spectral reflectivity of mixed dyeings can be calculated from the reflection data for dyeings with the separate components of the mixture.

Fastness of Dyed Fabrics to Dry Cleaning. Although it is generally considered that dry cleaning will not affect the color of dyed fabrics, no precise data have previously been available. Nearly a thousand dyeings of wool, silk, cotton, rayon, and other fabrics by representative dyestuffs have now been subjected to dry cleaning both with a moisture-free solvent and with a solvent containing one-tenth of 1