When the system was first introduced in Denmark, two methods of procedure were followed: the whole milk was heated to a sufficiently high temperature to thoroughly pasteurize it before it was separated, or it was separated first, and the cream pasteurized afterwards. In the latter case, it is necessary to heat the skim milk after separation to destroy the disease organisms, but this can be quickly done by the use of steam directly. Much more care must be used in heating the cream in order to prevent injury to the grain of the butter. In spite of the extra trouble of heating the cream and skim milk separately, this method has practically supplanted the single heating. With the continual spread of tuberculosis in America the heating of skim milk separately is beginning to be introduced.[169] ~Use of starters in pasteurized and unpasteurized cream.~ In order to secure the beneficial results presumably attributable to the use of a starter, natural as well as a pure culture, it should be employed in cream in which the bacteria have first been killed out by pasteurization. This is certainly the most logical and scientific method and is the way in which the process has been developed in Denmark. Here in this country, the use of pure cultures has been quite rapidly extended, but the system of heating the cream has been used in only a slight measure. The increased labor and expense incurred in pasteurizing the cream has naturally militated somewhat against the wide-spread use of the process, but doubtless the main factor has been the inability to secure as high a flavor where the cream was heated as in the unheated product. As the demands of the market change from a high, quick flavor to one that is somewhat milder but of better keeping quality, doubtless pasteurization of the cream will become more and more popular. That such a change is gradually occurring is already evident, although as yet only a small proportion of butter made in this country is now made in this way. Where the cream is unheated, a considerable number of species will be found, and even the addition of a pure culture, if that culture is of the lactic acid-producing species, will to some extent control the type of fermentation that occurs. Such would not be the case with a culture composed of the casein-digesting type of bacteria. Only those forms could thus be used which are especially well suited to development in raw cream. For this reason the pure culture ferments that are generally employed in creamery practice are organisms of the lactic acid type, able to grow rapidly in cream and produce a pure cream flavor in the butter.