_4. Elimination of taints._ Many defective conditions in butter are attributable to the growth of undesirable bacteria in the cream that result in the formation of "off" flavors and taints. If cream is pasteurized, thereby destroying these organisms, then ripened with pure ferments, it is generally possible to eliminate the abnormal conditions.[164] Taints may also be present in cream due to direct absorption from the cow or through exposure to foul odors.[165] Troubles of this sort may thus be carried over to the butter. This is particularly true in regions where leeks and wild onions abound, as in some of the Atlantic States. The heating of the cream tends to expel these volatile taints, so that a fairly good article of butter can be made from what would otherwise be a relatively worthless product. ~Characteristics desired in culture starters.~ Certain conditions as the following are desirable in starters made from pure cultures: 1. Vigorous growth in milk at ordinary ripening temperatures. 2. Ability to form acid so as to facilitate churning and increase the yield of butter. 3. Able to produce a clean flavor and desirable aroma. 4. Impart a good keeping quality to butter. 5. Not easily modified in its flavor-producing qualities by artificial cultivation. These different conditions are difficult to attain, for the reason that some of them seem to be in part incompatible. Weigmann[166] found that a good aroma was generally an evanescent property, and therefore opposed to good keeping quality. Conn has shown that the functions of acid-formation, flavor and aroma production are not necessarily related, and therefore the chances of finding a single organism that possesses all the desirable attributes are not very good. In all probability no one germ possesses all of these desirable qualities, but natural ripening is the resultant of the action of several forms.[167] This idea has led to the attempt at mixing selected organisms that have been chosen on account of certain favorable characteristics which they might possess. The difficulty of maintaining such a composite culture in its correct proportions when it is propagated in the creamery is seemingly well nigh insuperable, as one organism is very apt to develop more or less rapidly than the other. A very satisfactory way in which these cultures are marketed is to mix the bacterial growth with some sterile, inert, dry substance. This is the method used in most of the Danish cultures. In this country, some of the more prominent cultures employed are marketed in a liquid form.