30 BULLETIN 1265. U. S. DEPARTMENT OF AGRICULTURE 
driving air out of the tissues, and gelatinization of the starch, which 
results in a more transparent condition. Exactly what changes, if 
any, are brought about in the carotinoids, or other pigments, is not 
known. 
EFFECT OF PRECOOKING ON THE TASTE AND FLAVOR OF SWEET POTATOES. 
Sweet potatoes precooked in water and in steam showed no per- 
ceptible differences in flavor. Those canned whole in tin without pre- 
cooking sometimes developed an undesirable taste due to interaction 
of the metals and air on the canned material. The product from 
the potatoes canned raw, after passing through the food grinder, 
was likewise inferior in flavor to that canned after precooking. 
• CHEMICAL CHANGES IN SWEET POTATOES DURING PRECOOKING. 
In Table 5 are shown results of chemical analyses of the varieties 
of sweet potatoes sampled both before and after precooking for 35 
minutes in the steam chamber. The Nancy Hall represents the soft 
moist varieties, and the Big-Stem Jersey the firm dry varieties. Both 
were analyzed immediately after digging, and again after 30 days' 
curing and storage. 10 
In the publication already mentioned (39). the writers pointed out 
that the firmness or softness of the different varieties depends on 
the nature of the polysaccharide content. In the moist soft varieties, 
there is a high dextrin content and a somewhat larger amount of 
sugar than in the dry firm group. Owing to the difficulty of making 
complete separations of starch, dextrin, and sugars, the writers 
pointed out that the figures given for dextrin were probably much 
too low. It is shown in these pages that it is during the cooking 
that the starch is changed to dextrin, sugar, and other intermediate 
products. 
The samples for chemical analysis in the present experiments were 
prepared by passing about 2 kilograms of the precooked material 
through a food grinder, mixing thoroughly, and weighing a quantity 
into a large beaker. Enough 95 per cent alcohol was then added 
in quantity to make 80 per cent strength. The raw samples were 
heated in the alcohol to about 70° C, to stop enzyme action. After 
standing over night, the alcohol was decanted through an extraction 
thimble, and more 95 per cent alcohol added to the material. After 
thorough stirring, this alcohol was decanted through the extraction 
thimble, and the operation repeated four or five times. The residue 
was then transferred to the extraction thimble, and extracted in a 
Soxhlet apparatus, to remove the last traces of soluble substance. 
Sugars were determined by the copper-reduction method recom- 
mended by the Association of Official Agricultural Chemists (3). 
Determinations also were made on the alcoholic extract, both before 
and after inversion, as for cane sugar, and also after inversion 
with diastase. 
An attempt was made to determine dextrin, soluble starch, and 
insoluble starch in the residue after alcoholic extraction. The care- 
fully dried residue was ground to a very fine powder, and a weighed 
quantity extracted with water at room temperature. This was 
filtered through a Gooch crucible with asbestos mat. the residue 
• footnote, p. 28. 
