Starches isolated from 23 bread wheats (Triticum aestivurn) and 26 durum wheats (T. dururn) contained 26.3-30.6% (mean 29.1%) total amylose, 19.3-25.1% (mean 22.9%) apparent amylose and 783-1144mg 100gp' (mean 977mg 1OOg-l) lysophospholipids. Gelatinisation temperatures were 57.344.9"C (mean 61.8"C) and enthalpies 6 . 4 1 1 . 8 Jg-' (mean 9.7 Jg-') in excess water, measured by differential scanning calorimetry. There were no correlations between any of these parameters. Starch granule size distributions were determined with a Coulter Counter and 100-channel analyser. A-granule mean volumes were 1235-2.585pm3 (av. 1778), modal volumes 863-1804,um3 (av. 1264), mean diameters 13.9-16.0pm (av. 13.99), and specific surface areas 0.2360.302 m2 gp'. B-granule mean volumes were 35.4-100.4pm3 (av. 5.5.9), modal volumes 16.5-54.5pm3 (av. 27.7), mean diameters 3.665.07pm (av. 4.09), and specific surface areas 0.684-0.920 m2 g-'. The B-granule contents of the starches were 12.8-34.6% (av. 27.3) by weight (sedimentation method) and 13.@37.3% (av. 24.0) by volume (Coulter method), the latter being the more accurate method.

Starch was isolated from six types o f wheat and separated into large A-granule and small B-granule fractions which exhibited a suitable range of properties. Respective amylose contents were 28.6-31 .(I% and 25.1-31 .S%, lysophospholipids 707-1014mg 100g-' and 105&1398rng lOOg-', and gelatinisation temperatures (GT) by differential scanning calorimetry (DSC) 58.4-62.7"C and 60.5-64.5"C. The A-and B-granule fractions were remixed in various proportions and reconstituted with freeze-dried gluten and freeze-dried water-solubles from flour to test the effects of changes in the properties of the total starch, A-granule or B-granule fractions, and the effects of various proportions of B-granules on the quality of bread. The specific volumes of the loaves were not affected by starch amylose or lipid content, but were significantly correlated with either the G T of the total starches or their A-granules. The optimum proportion of B-granules was 2 5 3 5 % by weight, but their G T had no effect on loaf specific volume. Staling changes were quantified from crumb compressibility measurements and the enthalpy of the endotherm for gelatinisation of retrograded amylopectin in stored breadcrumb, measured by DSC. Initial and limiting moduli of crumb firmness were correlated with loaf specific volume and starch GT. but rate and time constants calculated from the Avrami equation were independent. Rate and time constants and limiting enthalpy values calculated from the DSC results were independent of all other measured parameters.