BACTERIA IN RELATION TO BUTTER-MAKING.

Otago Witness, Issue 2565, 13 May 1903, Page 6

BACTERIA IN RELATION TO BUTTER-MAKING.

In a preceding article a resume was given of Dr H. W. Conn's investigations of the bacterial influences productive of various fermentations ot milk. His experiments, undertaken with the view to test the effect upon cream-ripening of the various kinds of bacteria found in milk and cream, greatly enhanced the value of his labours, for, because of the minuteness of the records and their evident trustworthiness, even scientists recognised them as immensely meritorious. Experiments previously had been largely confined to a few selected species of bacteria, and consequently we had very little knowledge in regard to the effect produced upon butter by the many different species of bacteria commonly found in milk and cream. To determine accurately tlie relation of flavour and aroma in butter to | bacterial growth, it appeared to Professor Conn to be necessary to experiment not with one, two, or with half a dozen species, as had been generally done by bacteriologists, but with as large a number of thp species of dairy bacteria as possible. It was known for some years that the flavour, the aroma, and the acid which are produced in cream during its ripening, and which give tlu peculiar character to the butter made therefrom, are due to the growth of bacterie in the cream. The real source, however, of the flavour is the cream itself, and the quality of the cream undoubtedly affects the character of the flavour. It has been found that ripening is practically always accompanied by a souring of the cream. It has been found also thafc good flavours are especially developed by the acid-forming species of bacteria. It has been assumed, therefore, that ths development of flavour and the development of acid are essentially identical, or at least necessarily associated. Some practical, as well as scientific, butter makers are teaching that one essential point to be aimed at in the cream-ripening is to cause the acid-producing organisms to grow rapidly in order to develop an acid and flavour beforo the other organisms have i chance to increase. The fact that the extent of the ripening is determined by the amount of acidity conveys the impression that the ripening and the souring are identical. Some years ago, however, Dr Conn averred that flavour is independent of acid production, and that, while many of the acidproducing species also produce changes in the cream which give rise to a good flavour, equally good flavours may be obtained by ' species of bacteria that produce no acid, and that some species of bacteria may produce acid in abundance without giving xise to the proper flavour. This conclusion was also reached by other bacteriologists. Storch, who first worked with pure bacteria cultures for cream-ripening, found some species producing acid, but not good flavour, iind the same results were reached by Weigmann. The relation of ripening to the aroma of butter is also an uncertain fine. £ litre are several pure cultures usfd in different dairying countries for artificially ripening cream, most of which jjroduce favourable results so far as concerns acids and flavours, but none of which appears to give a satisfactory aroma. Each of these three factors seems to be essential ti a proper -cream-ripening, and we cannot hope to satisfactorily control this ripening until we know how, and under what conditions, flavour, acid, and aroma are produced. For over two years Dr Conn was engaged in testing the effect upon cream-ripening of the various sptcies of bacteria found in milk and cream, f hese experiments and their results are summarised hereunder. | In his series of experiments, Di Conn's first task was to collect from reameries and dairies within reach of the Storrs Agricultural Experiment Station, Connec- j ticut, a large variety of bacteria. It was especially desirable to obtain those Jound in .'reamerie. luring the months of May , and June, inasmuch as these months were : commoniy characterised by the production of the best quality of butter. During May 1 and June if 1895 quite a number of visits were made to creameries in Connecticut. Some if them were visited two or three times, others nly once. om the cream 4 Jius obtained "S many diiierent species of

bacteria as possible were taken at once, and set aside for future work. At subsequent periods other visits were made to the same places. Other samples of milk were obtained from dairies at Storrs, and from different dairies at a distance. From ; these various sources nearly 100 different ' types of bacteria *vere obtained, most of which were carefully studied and tested in i cream-ripening. All the general types of milk bacteria were included among this ; list. It included some bacteria which sour i milk by producing lactic acid, others which • curdle milk by producing a rennet-like^ ferm ent, but rendering the milk alkaline ; ' otherp, again, which exert a putrefactive effect upon the milk ; and still others that have seemingly no effect upon the milk or cream. The various types were in almost i equal abundance among the species coli lected, except that the number of forms that have no appreciable effect upon milk was considerably larger than those belong- ! ing to any >f the other classes. l In the early summer the variety of bac- ■ teria in the cream was found to be greater than at the other seasons. In nearly all the samples of cream collected in May, and particularly in June, the number of different species was very great, not only when different samples were compared with each other, but in the same sample of cream. j This would naturally have been anticipated, and is probably closely associated with the green food of the cows. It appears not unI likely that in this fact lies the explanation of the high quality of butter flavour commonly developed during these months. Not only is the variety greater, but the number of bactfci':a in the cream during these months is vastly in excess of that found under similar conditions in the cooler months ot the year. Another point in the same connection is the difference found at the same creamery at different times. Sucb samples, even though following each other at short interval, showed a considerable difference in the types of bacteria found. ! This is in part due to the fact that no bac- ! teriological examination of cream an dis1 close all of the kinds of bacteria therein, and the bacteriological analysis is, there- , fore, in every case very incomplete. Two 1 samples of the sanvj cream would doubtless ' show some difference for this reason. But this is not wholly the explanation of the matter, since it was frequently found that a sample of cream taken at one date would I disclose a large number of bacteria which 1 liquefied gelatine, while anothei taken a few lays later would show no liquefying bacteria. The presence of tlu liquefying organisms is mosc easily detected ; in fact, one can never fail to detect them. Their presence in quantitr in some cases and their absence in others is therefore significant. j Variation with the Cow. — One series of experiments consisted in the testing of the milk from eight cows in th<» same barn. These cows were kept in adjoining stalls and fed in the ssme manner, and their milk was drawn into sterilised bottles and then tested separately. After a .'ew weeks the same eight cows were again tested in the same way, and the same test was repeated j at short intervals for several months. It was found in these tests that there was ! the most striking difference between the bacteria in the milk of the separate :<i\vs. j The number varied surprisingly. The milk from two of the cows contained not more j than 250 bacteria per loop full (a loop full is a drop about the siz-s of the head of a i large pin), while the milk from a third contained 20,000. :md a fourth 60.000 in the same quantity of milk. The variety of bacteria was no less interesting. In the first place, it was found that no two of , these eight samples r.f milk, when left to themselves and careful!}' guarded fiom out- i side contamination, underwent the same ! kind of fermentation. Some curdled and | soured, some of them curdled without sour- : ing, some developed a cheesy odour, others j a putrefactive od'iur, and among the lot j there was one ow that gave milk that became slimy. When the same cows were tested a second time, a /ew weeks later, the effect were different. The cow that pave slimy milk no longer produced milk with this defect, and all >f the samples, but one soured, although m>t in the same way in any two cases. In the third test still othc-r variations occurred. When the bacteria from the eight samples were studied it was found, as wag to ba expected, that there wns a good deal of variation. There were one or two species that were common to nearly all of the cows, while others were found in one lot of milk, and others, again.

in one or two lots. It must be always kept in mind that these bacteria did not j come from within the milk gland, but only ! from the milk ducts. They do not therefore come from within the animal, but really from the exterior. They are bacteria from external sources, which have made their way into the ducts, and not bacteria from within the animal, making their way out. This difference in the bacterial flora of milk from cows in the same barn is certainly a surprising and interesting fact. It gives us a suggestion as to the complex mixture of bacteria in cream of an ordinary creamery coming from hundreds of cows. It shows, further, how impossible it must be to obtain a uniform quality ot cream (so far as bacteria are concerned) from many contributing sources. METHOD OF EXPERIMENT. The method of experiment was to separate a lot of cream from the milk by a centrifugal process, and then divide it into equal parts. Ail of the cream was treated co a temperature of 156-158deg Fahr. for 15 minutes, and then allowed to cool. This heating (pasteurising) destroyed most of: the bacteria which chanced to be present in the cream, only such .bacteria as produce spores remaining alive. Experience has shown that such heating will kill all the lactic bacteria. The species of bacterium to be tested was grown in sterilised milk. Two days before the experiment began a sufficient number of vessels of sterilised milk were inoculated, each with a different species of bacterium. These were then allowed to grow for two days. When the lo^s of cream were pasteurised and cooled as above described, one of these milk starters was poured into each. Each of the lot received a starter made from •>. different species of bacterium, and one lot was always left ; for a control experiment without any I starter. ; The samples were then placed under similar conditions as to temperature, an.l allowed to ripen for the same length of time. After considerable experience it was found that the most satisfactory method of procedure was to use a ripening of 48 hours at a somewhat high temperature (about 70deg Fahr.). After the ripening the cream was cooled and churned. The examination of the butter was made in most cases without salting, inasmuch as salting very commonly obscures the peculiar flavours developed during the ripening process. The testing was made for. flavour, for acid, and for aroma. Inasmuch as the experiments were designed to test the general effects of many species of bacteria rather tJian -to find "out the particular species which produced the best butter. Professor Conn thought that this method of testing several species simultaneously promised the most valuable results. RESULTS. Control Cream Compared with Inoculated j Cream. — One of th. most interesting facts J was found in comparing the control (i.e., j the pasteurised but not. inoculated) samples I of cream, and butter made therefrom, with ■ the inoculated samples. As a rule the control butter possessed neither flavour nor aroma — in no case unless the ripening had continued too long. Nevertheless, it was found in many cases that the control cream did undergo some decided changes during the period of ripening. The temperature of 158deg Fahr. (used in pasteurising) does not kill all the bacteria in the :ream, and the subsequent ripening being somewhat long and the temperature somewhat biiii, the few bacteria that were left in the creaii after pasteurisation had time to develop. The cieam thus frequently showed the effects of their presence. In many cases the control cream was thick and nearly curdled, but inasmuch as it was never acid, it was plain Jhat this effect was due, not to the lactic acid organisms, but rather to the growth of the species of bacteria which curdle milk by production of a rennet ferment ; a class frequently called the putrefactive class of bacteria. This is readily understood, since these bacteria frequently produce spores which resist heat, while the acid bacteria produce no spores. In a few cases the control cream became sligntly bitter, or developed some other unusual taste, but the taste was so slight that it had no effect upon the butter made froni tiie cream. These facts, of course, aye not surprising, for they are exactly what might be expected when we remember that pasj teurisation does not destroy all the bacteria | present in tho cream. The interesting , fact in these experiments was that in no ; case did the inoculated lots of cream show similar results. In one case, where the , control :ream became bitter, none of the I three inoculated samples showed the slightest trace of bitterness. Where the control cream showed a partial curdling, the inoculated samples showed an entirely independent effect, that was evidently due directly to the influence >f the inoculated species, and not to those left in the cream niter pasteurising. In some cases the inoculated cream was thickened and curdled from ' the effects of he bacteria with which it had b?en inoculated, but in other oases, where the inoculated species had no power of curdling the cream, the cream at the end of the ripening was as thin as at the beginning, showing no trace of curdling, even though the control cream was at the same time" very thick. These results were not obtained in one or two cases, but m a great number of experiments. The result it first surprised Dr Conn, but it ""as found to be so general that he . c oon :an;e to look upon it as normal, and to expect it. WHY STARTERS ARE BENEFICIAL. The importance and significance of tl'is fact is considerable. If the control levrlops a bitter taste, while the inoculated species dots not, this can only be because certain bacteria grow in the control which do not grow tc an equal extent in the inoculated cream. When ream is inoculated with one kind of bacteria in considerable quantity, other species of bacteria already present may be checked in their development by the growth of the inoculated species and prevented from producing their normal results. The control and the

inoculated cream must have had at the end of pasteurisation the same kind of bacteria present, but the inoculation of the cream with >ne species of bacteria in the artificial culture prevented those in the cream either from growing or from having . their normal effect upon the cream. This result is, indeed, not very surprising after all. Bacteriologists have for some time known that different species of baci teria may thus have a repressing influence ' upon each, other. It 'has been deter* mined, for instance, that the growth of the normal bacteria in milk prevents the growth ; there of the cholera bacillus, ftlthougb. the cholera bacillus will grow readily in milk , that has been sterilised. Many other similar instances have been found, indicating that this is not an unusual, but rather a common effect where different kinds of bacteria are growing side by side. The importance of this matter to the buttermaker is considerable, inasmuch as it indi- | cates that it may be possible, by lnoculatj ing cream quite heavily with one kind of i bacteria, to check the influence of the other I kinds which may be present. One can thus obtain the influence of the inoculated species, but little modified by the growth of the other bacteria which ire present in less abundance. During the last year or two butter makers have become convinced of the advantage of using starters. They have found that in many cases the use of a starter — either a natural starter or one of the various pure cultures which are on the market — will improve their butter where it is added to ordinary cream. It has been something of a question how j starters can do any good in cream already j more or less in^n-egnated with bacteria. But if, as these experiments show, such a starter has the power of checking the growth of normal bacteria, we can understand the matter. If starters can have any influence in checking the of bacteria .already present, we should expect that such starters would frequently improve butter, although not always. Thus the facts here given offer an explanation of and emphasise the value of a starter of some kind in cream, both for the purpose of starting the proper kind of ripening and also to check the development of many bacteria already present which might be injurious to the butter. MOST BACTERIA HARMLESS OR BENEFICIAL. The majority of the species of bacteria tested by him, Dr Conn says, may be regarded .as indifferent in their effect upon the j butter. About half of them when used to ripen the cream produced butter that had neither flavour, nor aroma, nor acid, and. the butter was practically indistinguish- : able from the control butter. These species are in the greatest variety, and largest in number around cow barns and dairies. They are perfectly wholesome in uhe cream. They do no injury, they do no special good, and it may, therefore, be concluded thati the majority of the species of bacteria that are present in the sources of our milk are wholesome forms. This, of course, where all the surroundings of the premises are cleanly kept. A considerable number of the species are positively favourable in their influence on the butter. Of the 68 species tested, 20 produced Dutter described in the notes as good-flavoured. Of course the flavour was somewhat variable, and its good character, while sometimes striking, was at other times moderate. It was not always the typical butter flavour, and yet was such an approximation towards it that the butter would be regarded as of a good quality. A smaller number of species produced injurious effects upon the butter. Eighteen species among the 68 tested were described ,s producing butter that was bad, or poor, or strong-flavoured, or disagreeable ; various adjectives being used to indicate the different effects. Sometimes the poor flavour was a putrefactive taste, "while in other instances the effect Mas of a peculiar, indescribable character. In many of these 18 species the unpleasant flavour was very slight, and probably insufficient to materially injure the butter. FLAVOUR INDEPENDENT W ACID. Of the species of bacteria producing good flavours in butter many were of the acidproducing class. Of , the . 20 above-men-tioned, nine were lactic organisms.'- On the other hand, 11 were among the class which would be described as alkaline species, by which it is meant that they either produced an alkaline reaction in the milk or produced no :liange in its reaction. They are, at all events, distinctly not acid form?. Seven among them liquefy gelatine, and are, therefore, among what are called the putrefactive bacteria. In speaking of flavour, it must be distinguished from acid taste. Of the 18 species described as producing injurious effects upon the butter, nine belonged to tha acid-producing class, while nine belonged to the class developing alkaline reaction. From these facts Dr Conn regards it as a safe and perfectly legitimate inference that flavour is a matter entirely distinct /rom acid. It will be noticed that among the species producing good flavours in the butter, while many of them are acid producers, a large number (11 out of 20) are among those that develop no acid. Ifc may commonly happen, as will be noticed from these results, that the same species of bacteria mar develop acid and flavour. This undoubtedly is the case with many of. the bacteria of milk, and with most of the species of bacteria that are used by various butter makers as cultures for artificial fermentation. Whila the lactic bacteria may be regarded as commonly producing the butter flavour in practical butter-making, they do not do this simply because theyi produce acid, and it must be recognised that other types of bacteria assist in producing the desired flavour. It is important;

Eai«rs of Poultry should use ITntvo AKD Blair's Game and Poultry Meal, which is composed of the best ground bones and shells, specially prepared to meet the requirements of the last-increasing poultry industry. It i«made up in J-llb bags, at Is 3d each. Ask your storekeeper fci it.

-to note in this connection that, of the 30 species described as indifferent in their faction, none were acid organisms. AROMA INDEPENDENT OF FLAVOUR AND ACID. Perhaps the most interesting result has to do with the production of the butter aroma. The butter aroma, the character that affects the nose rather than the palate, but a sure test of quality, appears to be — at least-, so far as the results of the experiments are concerned — entirely independent of the flavour. Moreover, it appears to be a more unusual thing for bacteria to produce a desirable aroma than a desirable flavour. The great majority of the species tested gave rise td practically none, or at least to an extremely slight aroma. Of 65 species whose effect on aroma was observed, 39 produced no aroma at all. Of the species of bacteria which thus bave no influence upon the aroma of butter, the majority, again, are among the class which either develop an alkaline reaction in the cream or do not change its reaction at all. Seven of those producing no aroma are among the class that produce lactic acid. "Among j those that do produce an aroma, of a de- j cided character, 18 are described in my i laboratory notes as producing an unpleasant ; .or a bad aroma; seven of these are among .those that produce lactic acid. The kind .of aroma developed varied widely in these different species. Sometimes it was an extremely sour smell, at other tijnes it was in a measure putrefactive. In most cases the aroma was of a character that was indescribable, from the lack of proper, terms, but always unpleasant, and would always "be regarded as characterising a poor quality of butter. Among the 60 species studied, only eight have been found as yet to produce an aroma which has been described in my notes as good ; and in only three has the aroma been that which is looked for in first-class butter. In two or three case 3 the aroma produced was of an extremely fine character, and in these artificial tests almost identical with the aroma expected in the first-class butter from a creamery. It has been interesting ta find that, of the eight species which produce the aroma which has been described as good, none have been among the acid-producing organisms. The eight either develop an alkaline reaction or have no special effect upon the reaction of milk. There were three which developed tbo most typical aroma of alj the species studied (Nos. '66, 69, and 104). Two of these curdled milk by producing a rennet, both liquefying gelatine. The third did not curdle the milk." The results surprised Dr Conn, who had supposed before the experiments began that the aroma was a matter very closely associated with the j development oi the 'actic acid.