The basic features of staple yarn structure which can vary in cotton yarns include those which depend on fiber factors such as fiber fineness, maturity, and staple length. Then there are mechanical processing factors, such as rate of carding, combing, and singeing, which affect fiber arrangement in or on the surface of the yarn. Finally, there are whole-yarn factors, such as twist, plying, and yarn fineness, which affect luster. These are discussed in three sections, which are presented together because of the use of experimental material which compares more than one type of factor at the same time. The results show that the three main means of improving luster by yarn structure factors are: ( 1) lowering of twist; ( 2) use of optimum plying twist, as shown by Adderley and Oxley ; ( 3) singeing to remove free surface fibers. Two additional structure factors, which, however, will not usually produce as great an effect on luster at the same cost, are: ( 4) using finer yarns; ( 5) combing.

The flattened, twisted-ribbon shape of the natural cotton fiber is unique among textile fibers, and leads to characteristic differences in the reflection of light from cotton, in comparison with that from more rounded fibers. With natural cotton, there are a large number of small planes set at every possible angle, so that light is not merely scattered, but also is in part reflected in a regular mirror-like way at every angle. However, small planes which are set at the more glancing angles reflect a larger proportion of the light which falls on them, in a manner described quantitatively by Fresnel's law. This explains the shift of the peak of reflection to angles higher than the mirror angle for the general surface, a shift which is greater for cotton than for most other fibers. If corrections based on Fresnel's law are applied, the peak of corrected reflectance is found to be at the mirror angle for cotton.The relations between luster and the balance between regular reflection and scattering are also shown in the higher luster of sheer fabrics and the decreased luster caused by increased scattering from deeper layers. Another demonstration of the favorable effect of decreased scattering on luster is shown by the increased luster of dark colors, where absorption prevents scattering and leaves the field dominated by surface reflection.In fabrics, there is scattering by the crossways system of yarns. If the Fresnel corrections are applied, the effect of the crossways yarns can be largely removed, and the effect of weaving crimp made clearer. Weaving crimp produces secondary or shifted peaks of reflection close to and on each side of the main peak, which correspond to the yarn rising up, going over the cross-yarn, and dipping down again.

Luster is not an "absolute" property of the cotton or the yarn, but depends on the sheerness of the fabric, as well as on the weave pattern. One of the reasons for the greater apparent luster of sheer fabrics or thin layers of yarn is that there is less opportunity for light to be scattered within thin layers than in its return from the lower parts of thick layers of yarn. Such scattering adds an approximately equal amount to the reflectances at all angles and so lowers the apparent contrast ratio for the thicker layers.At the opposite extreme, the reflectance of individual yarns can be determined, and it is demonstrated that yarn removed from the fabric has nearly the same luster as the original yarn, if the weaving crimp is removed. Weaving crimp by itself reduces the measured luster, tending in the same direction as the effect of cross-yarns.

The entire range of American cotton production has been examined for correlation of luster with other fiber properties over two crop years. This has been possible through the cooperation of the Division of Cotton and other Fiber Crops of the Bureau of Plant Industry, U.S.D.A. which provided the yarns and the fiber data.Luster is shown to be a varietal characteristic, although also influenced by environmental factors. The strongest correlations are of higher luster with greater length, greater strength, and greater roundness. American-Egyptian types, experimental triple hybrids, and Hopi Acala 50 are each different in the relation of fiber properties to luster. All the rest of the cottons, that is, the main Upland group, whether rain grown or irrigated, or long staple, show a uniform general trend of relationships. This trend of relations offers a guide to both spinners and breeders who may wish to select cotton for greater luster. The selections for luster which are possible within the main crop amount to as much gain in luster as can be obtained by plying.These results are mainly for natural or raw cotton, but the trends remain generally the same after mercerizing, with a tendency for the most lustrous cottons in the natural state to be the ones which gain most on mercerizing.

The relations of luster and fiber properties have been studied in numerous specimens of a limited number of varieties, both as natural and as mercerized yarns. The differences of group averages for natural yarns of several Upland varieties appear due to yarn structure factors, but the high luster of American-Egyption cotton must be due to other factors. The gain in luster on mercerization tends to be greater, the greater the original luster, both between and within variety groups. Within one variety, Deltapine 15, the trend of the luster of mercerized yarn is to increase with increase of the fiber properties of length, wt./in., and maturity, or combinations of these. Fiber strength has no consistent relation to luster within this group. Length shows the same between and within group relations, but wt./in. shows the opposite relations between groups, and maturity shows no trend between groups. Fiber properties or combinations can be used to select for better than average luster in the longer cottons, but there is still considerable range and overlapping. The improvement possi ble by selection within a group is small compared with either mercerization or use of American- Egyption instead of Upland cotton, but suggests the worth of continued studies in this direction.