Minard and Myers both found that the texture of the fat is important in relation to creaming quality and in obtaining velvety texture in cakes. By the rate of cooling a fat after it is melted, its texture can be controlled so that it is very waxy and firm, soft and smooth, or granular with a lot of separated oil. Minard found the very firm, waxy lard did not cream well. Myers found a soft, smooth lard creamed best, butter ranked second, and grainy or granular lard was poorest.

Since any fat tends to become harder at low temperatures and softer at higher temperatures, there is a range at which each fat creams best. Martin used hydrogenated lard, Crisco, 44 lards of varying softness and firmness, and butter. The initial temperature of all the fats at start of creaming was 24° to 25°C. All creamed fairly well at this temperature, though in general they tended to cream better at higher temperatures. In general the volume was decreasingly smaller in the order named. A low initial room temperature retarded creaming. Creaming tends to elevate the temperature of the creaming mass, probably from friction, though the butter usually shows an initial drop in temperature. For optimum creaming the firmer fats require a higher temperature than 25°C. The hydrogenated lard creams well at 30°C. or above this temperature, creaming the best of any fat tried on hot days. Optimum creaming temperatures for fats commonly used are 24° to 29°C. (about 75° to 85°F.). In the studies of fats in relation to their use in cake at Iowa State College a highly significant correlation between the volume of the creamed mixture and their temperature has been obtained, which indicates the volume of the creamed mixture is related to its temperature, as expected.

Butter varies in creaming ability. Bakers usually report that it gives a poor creaming volume. Jenkins used three winter butters with Crisco as a control. She used Formula I with 8.4 grams of a tartrate baking powder for the full recipe. The butters were prepared in January by the Dairy Department. The same batch of cream was used for the three butters, but portions were churned when the serum acidity was 0.2, 0.4, and 0.6 per cent. This gave butters with acid numbers of 1.20, 1.40, and 1.70, respectively. The acid number of the Crisco was 0.10. The creaming volume was least from the butter of intermediate acidity, greatest from the most acid butter, and the other two fats gave practically the same volume. It is rather interesting that, regardless of creaming volume, the final volume of all the cakes increased with an increase in temperature of mixing. The final average volumes of the cakes were very similar. But the fat giving the greatest creaming volume does not seem always to yield the greatest cake volume. The temperature of the ingredients when mixed had more effect on the final volume than did the creamed volume of the fat. No indication of the effect of variation in composition of butter on creaming volume was obtained. The composition of butter varies with the season of the year and with the locality in which it is produced. Graham using summer butters obtained results that were very similar to those of Jenkins, but the cakes had smaller volumes. It may seem rather strange that there is no close correlation between the volume of the creamed mixture and the volume of the cake. But studies tend to show this. Both Martin and Buel obtained nonsignificant correlations between these two factors. Martin obtained a negative correlation which, although not significant, tended to show that, for

4 fats out of 5, smaller cake volumes were obtained with larger creaming volumes. Minard obtained significant, negative correlation for these two factors showing that in her studies cake volume was smaller when the creamed volume was large. The fats studied in these investigations have included several types of lard, butter, hydrogenated lard, and Crisco, though the studies with the last fat were not extensive. From her results, Martin suggested that there may be for each fat a point short of maximum creaming volume, at which the creaming process should be stopped in order to produce a cake of maximum volume.

But what Martin calls creaming quality is important and is closely related to velvety texture and high scores for cakes. By a good creaming quality is meant that a soft, spongy, not too frothy, not a granular, lumpy, or pasty texture is obtained. In other words, there is a particular body and texture in the creamed mass rather difficult to describe that is related to excellent textures in cake.

Sugar. Granulated sugar is most commonly used. Finely granulated sugars dissolve more readily than coarse ones. Minton states that, the finer the granulation size, provided it is a natural granulation obtained in the refinery and not a powdered or crushed sugar, the better the creaming quality. The Bakery Research Department of Procter & Gamble Company also state that the fine crystal that has a glossy surface, and is not made by grinding or crushing, gives better creaming volume than the coarse sugars. Brown sugar, because of its flavor, is often used in cake. It is preferable to substitute it by weight rather than by measure.

The transformed sugar, with its irregular surface which holds much air, should not only carry air into the batter, but many films of fat.

Liquid. Water or fruit juices may be used for the liquid in cake. When water is substituted for an equal weight of milk, a more moist cake is obtained. Milk contains about 12 to 13 per cent solids.

Baking powder. Patterson states that unless the cake cells are very fine and uniform a very fine baking powder should be used, so that small bubbles are formed. With many large bubbles of gas formed at the same time, the cell walls may be subjected to too great a pressure and collapse, producing a fallen cake.

Any of the different types of baking powder can be used in cakes. The main discussion of the use of baking powder in cake is given later in the chapter under "Variation of Proportion of Ingredients."