CHESE YIELD INVESTIGATED

Taranaki Daily News, 16 July 1929, Page 14

CHESE YIELD INVESTIGATED

’SCIENTIST’S EXHAUSTIVE REPORT

GUIDANCE FOR THE MANUFACTURERS

INFLUENCE UPON FACTORY PAY-OUT

Continuing his report to the Federation of Taranaki Co-opeia-tive Dairy Companies on the question of cheese yields, Mr. P. 0. Veale, scientist in charge of the federation’s laboratory, exhaustively discusses the effect of moisture content on yield, the difference between raw and pasteurised milk, the effect of curing room conditions, variations in yield'produced by part-skimming, factors (influencing false or theoretical calculations of yield, the [effect of part-skimming and 11 reading down” the test, and the influence of yield upon the pay-out on butter-fat.

After a discussion of the report by members of the Factory Managers’ Association the following resolution was carried:—“This meeting wishes to record its appreciation of the opportunity of discussing with Mr. Veale and departmental officers the paper submitted by Mr. Veale on cheese yields, and also wishes to record its agreement with the views expressed therein.”

MOISTURE CONTENT OF CHEESE. The moisture content of cheese is well blown to play a most important part in letermining yield —so well known, in fact, that some persons are 'inclined to attribute all the differences in yield between factories to this cause, the report Continues.

■ Up to. a certain point, the moisture lonteni of cheese is the main and controlling factor in determining yield. Lowtest milk with a high casein ratio _ certainly contains more casein per unit of putter-fat than high-testing milk, but the chief reason for its high yielding rapacity is the ability of the casein to Bold moisture without appreciable loftening of body. A , typical cheese M this variety may show the following Compositions: Fat 33 per cent., moisture ES per cent., and yet have equally good body and texture as another cheese made from high-testing milk, having an analysis; Fait 37 per cent., moisture 33.5 per pent. Each pound of. butter-fat from low-testing milk therefore is associated with more casein and much more moisture than is the case with the product bf high-testing milk, consequently yields bf ’over 2.701bs of cheese per pound of putter-fat are quite common from such milk, without impairing in the slightest rhe body and texture of the resulting bheese.

It is not possible, however, to apply ;his principle outside of its own proper sphere, and to obtain high yields of good ffiees.e ■by artificially raising the mois;ure content. If an attempt is made to ’aise the percentage of moisture in :heese appreciably, above the natural figire dictated by the casein.ratio of the iriginal milk’, weakness of body imnediately results, and the setting up of mdesirable fermentations is facilitated, fhese faults are discernible by our gradng officers, and erring cheese-makers are speedily advised. This knowledge has >een forcibly driven home in Taranaki vithin the past few years, and chcesenakers now realise the futility of enleavouring to raise their returns by exerting water. 1 ■ EXCESSIVE MOISTURE.

Of course, good practice and common sense dictate that a cheese should contain the full moisture content consistent with quality, and managers are Only doing their duty in attending to the matter, for lowered yields will undoubtedly be experienced if an unduly dry cheese is made. Unfortunately the notion is still prevalent that New Zealand cheese-makers in general and Taranaki cheese-makers in particular, are continuing to strive after inflated yields by incorporating excessive moisture, and that this practice is responsible for the. national fault of openness in our cheese. As a result of thousands of analyses of South Taranaki cheese made over the last four years, it. may be stated with confidence that this contention has no, foundation .in fact. It has not been found possible to demonstrate a consistently high moisture content in the cheese of those factories which publish a high, yield (i.e. substantially above 2.65 from a 4.0 average test), whereas, bn the contrary, low moisture percentages have very frequently been found vrith some factories whose published yield figure is low. These facts will become evident from a consideration of the following analyses of cheese from two different factories which finished almost at opposite ends of the field in. the “race for yield” during 1927-28. All figures are averages for the number of samples indicated, the cheese being fiom 10 to 14 days old.

In the first place, the moisture figures for Factory A are by no means excessive, thus proving that their high yield was not obtained by the incorporation of excessive moisture. Van Slyke and Price have laid it down that cheese should reach its consumers containing not less than 35 per cent, of moisture. It is quite obvious that Factory A has kept well within the mark. Of course, it may be argued that Van Slyke’s standard refers to American conditions, and that New Zealand conditions may be such that even 35 to 36 per cent, of moisture is excessive. As against this, we may quote the analyses of 46 samples of prize show cheese examined during the years 1024-26 bv Professor Berry at the ;West of Scotland ■•Agricultural College. The lowest average moisture content was C S3.I in 1926. and the highest average was 34.5 in I9W. Tb" r’-nnc-e o f Factory A

would have the greatest difficulty in retaining even 33.1 of moisture by the time it reached England, thus proving that its moisture content could never be considered excessive, and might even be judged on the low side. Professor Berry also gives a comparison between 10 samples of New Zealand cheese and six of Canadian in 1927, showing that the average New Zealand moisture content was 32 9 per cent, and Canadian 32.5 per cent. The moisture figures already quoted for Factory A would easily have, been reduced to the vicinity of 32.5 to 33 per cent, by the time the cheese reached England. Hence it is fair to conclude that the high yield obtained by Factory A was not obtained through excessive moisture and further that openness in New Zealand cheese is not caused by an abnormally high moisture content, neither are our highest yielding, factories necessarily the chief offenders. The figures quoted for Factory B are nearly all lower than those for Factory A, thus showing that a low yield may be correlated to some extent with a reduced moisture content. It should be understood, of course, that the moisture must be considered in relation to the fat percentage, and that, if the fat. is high, the moisture cannot also be high without detriment to the body. As an example of this principle we may state that Factory A in January, 1928, has, in incorporating 35.7 per cent, of moisture with 35. G per cent, of fat, made a better performance than B, which incorporated 35.6 per cent, of moisture with only 34.8 per cent, of fat. Summarising the arguments under this section, we may state that only up to a certain point is the moisture content of cheese effective in determining yield. ;Low yields may be obtained where the 'moisture content of the cheese falls below normal, but, on the other hand, high yields (i.e., in the vicinity of 2.701bs of cheese per pound of butter-fat from a 4.0 test) are not obtained through the incorporation of excessive moisture when texture and body are known to be satisfactory.

RAW AND PASTEURISED MILK. In view of the fact that the large majority of New Zealand cheese is now manufactured from pasteurised milk, no discussion of yield would be complete without a consideration of the effect of pasteurisation. Practically no authoritative data are available from carefully controlled experiments using the same milk raw and pasteurised, and we must therefore fall back upon the fruits of experience. It is generally held that the pasteurising process results in a slightly increased yield of cheese because it allows the incorporation of slightly more moisture without detriment to body or flavour. The reasons for this may be discussed as follows: Firstly, the cheese-maker handling raw milk is practically at the mercy of the bacteria which inhabit the milk supply. These may produce gas, they may cause imperfect cooking, or they may inhibit the action of the starter. As a result, the imperfections of body are caused and these are frequently corrected by such methods as excessive dry stirring or heavy salting. Consequent upon these rather drastic methods of expelling moisture we find that raw milk cheese very often fails to realise its maximum potential moisture content. Even if deleterious bacteria are not present in large numbers, the cheese-maker is often unwilling to take the risk in allowing the curd its full complement of moisture, because ho realises that this .moisture will stimulate bacterial fermentation, and may thus facilitate the development of an undesirable flavour which would otherwise have failed to assert itself. On the other hand, the cheese-maker handling pasteurised milk is free, the majority ci his working season, from 51 the disturbing .influences of harmful micro-organisms. He can often time his operations strictly by the clock. He is free to give hL whole attention to the production of a. good body without having to sacrifice anything in order to obtain it and he can allow the curd its full complement of moisture, secure in the knowledge that fermentation will not later develop. A final point and probably the most important of all, is the fact now being established by experimental evidence that the process of heating causes some change in the constitution of the casein, whereby it is able to hold moisture more tenaciously. Hence even when finest quality - raw milk is handled, the yield of cheese is not as large as that from the same milk when pasteurised. It therefore works out in practice that pasteurising factories achieve . a somewhat better yield than raw milk factories, but the extent of the advantage is not precisely definable, as it naturally depends on the extent of the sacrifices which had to be made when handling the raw milk. As an indication, howevey,- we may quote the experience . of three factories which adopted pasteurisation during 1927-28, and showed for the season =an average rise in

yield of 0.061bs of cheese per lb of butter-fat. It was a year of slightly increased yields apparently, as their two nearest neighbours (non-pasteurising) also showed increases averaging.o.o2. Allowing this as due to seasonal influence, we still have approximately 0.041bs of cheese per lb of butter-fat which seems to be the result of the pasteurisation process. . , These particular factories, however, showed during the season evidence of slight over-moisture in their cheese, and, while this was occurring,. grades were not very satisfactory owing to . weak body. This indicates that the rise ,of 0.04 in the yield due to pasteurisation was somewhat abnormal, and might not be applicable to the average factory without a similar drop in quality. We may be justified therefore in not adopting the rise of 0.04 as a. standard, and, since a conservative estimate is desirable, 0.02 will be tentatively named as the rise in the yield which could reasonably and confidently be expected from the adoption of pasteurisation. Uns would modify the standards already quoted, making them approximately as follows: —

EFFECT OF CURING ROOM CONDITIONS UPON YIELD OF CHEESE.

During the 14 days which cheese stands upon the factory shelves before being weighed for export, it is constantly shrinking, the loss averaging approximately 2 per cent, over a whole season. For all ordinary curing conditions, this figure is sufficiently constant to be taken as a standard, but it may undergo important modification where either cool curing or early paraffining of the cheese is adopted. In consequence, yield figures may be slightly raised, as "the standards already quoted are based upon an average of at least 2 per cent, shrinkage from the date of manufacture. In the cool quring process a well-in-sulated curing room is employed, and by the aid of artificial refrigeration, the air temperature is kept below 50 degrees Farenheit. This possesses two advantages apart from the effect of the low temperature in hindering the development of any undesirable bacteria which might be present. These are: —(1) The rate of evaporation of the moisture from the cheese is lowered at the lower temperature; (2) through the necessity of maiijisuiing. a low temperature consistent with economy with the refrigerating plant, .doors are nearly always kept closed, thus raising the relative humidity inside-the curing room and still further reducing the rate of evaporation from the cheese.

Data are not available to show precisely the effect of cool curing upon yield, but a figure can be estimated from the known average reductions in shrinkage. From these it would appear that 0.021bs of cheese per lb of butter-fat should cover the improvement in yield consequent upon the saving in shrinkage as compared with a well-managed, ordinary curing room.

PARAFFINING. On the subject of paraffining, nothing can be given as a standard concerning, its effect in reducing shrinkage, as the time of paraffining has been so different in different companies. Some factories have claimed a saving in the past by early paraffining, but the scheme is falling into disfavour, as certain defects in body and rind have been accentuated by the process. Since it is now required by regulation to coat with paraffin and to pack at 14 days in the one series of operations, the procedure ceases to exercise any effect upon yield as computed from the factory packing weights. On the other hand, paraffining has an important influence if nett yields, which allow for the -shrinkage on the voyage Home, are under consideration. It is usual to allow 24 per cent, for this, but efficient paraffining has reduced the shrinkage to about per cent, or less. A few companies still publish nett yield figures in their balance sheets, but in the majority of cases, comparisons are made between factories on the basis of the gross yields from packing weights, and unless anything to the contrary is stated, gross yields are assumed.

VARIATIONS IN YIELD PRODUCED BY PART SKIMMING.

Since the days when competition in yield reached the acute stage, there has been the suggestion that sonic of the high yields published on annual balancesheets have been secured by a process of part-skimming, either by the use of abnormally large quantities of a skimmed starter, or by the illicit use of sweet skim-milk in the cheese vats. The high yield of cheese per pound of butter-fat obtainable from low-testing milk is well known, and the fact is cited in explanation of some of the yields claimed. The “explanation,’ however, is a fallacy, and it is proposed to show that partskimming docs not raise the factory returns for the yield of cheese per pound of butter-fat.

The way in which the fallacy arises can best be shown by an example: Consider lOOOlbs of average milk with a 4.4 butter-fat test. According to the standards quote this milk would give a yield-of 2.49 and would therefore give 109.Gibs of cheese (i.e, 2.49 x 441bs). Suppose now, that this milk had been skimmed down to a butter-fat test of approximately 4.0 per cent, by the removal of 41bs of Lutter-fat.' The yield per pound of butter-fat contained in it is now 2.61, and it would accordingly yield 104.41bs of cheese (i.e. 2.61 x 401bs). Estimated, however, with reference to the total butter-fat originally received and credited (this, being the method which the factories adopt) the yield per pound of butter-fat for the part skimming process becomes 104.4 (cheese) divided? by 44 (butter-fat) equals 2.371bs of cheese per pound butter-fat. In just the same way in the operations of a factory, part skimming lowers the total output of cheese because it is a removal of raw material; it raises the yield per pound of butter-fat calculated with reference to the butter-fat actually used in the vats but lowers the yield per pound of butter-fat calculated with reference to the total butter-fat received into the factory. It may not therefore be clear ’at first sight as to why part-skimming of fair-

ly rich milk is considered profitable. One may begin to answer this question by observing that, with normal prices ruling, the return from cheese-making is usually a trifle better than that from butter-making. In other words, the fact that cheese consists approximately of equal proportions of butter-fat, moisture, and casein, etc., is more than sufficient to counter-balance the superior price per pound for butter-fat alone. Now, when rich milk is part skimmed, a quantity of butter-fat is made available for sale at butter prices, while the yield of cheese from the remaining milk is reduced by not a great deal more than the actual weight of butter-fat removed. The yield reduction is never as large as the cheese-equivalent of the butter-fat removed, because the casein is practically unaffected in quantity, and the moisture content of the part-skimmed cheese, although slightly less in actual weight, bears a higher proportion to each unit of butter-fat in the cheese.

Hence the process of part-skimming allows us to sell in the resulting cheese, butter-fat associated with a higher proportion of casein and a higher proportion of moisture than before, without detriment and often with considerable improvement to body and texture. This is the real source of the profit from reducing the fat percentage of very rich milk for cheese-making. Without this superior moisture-carrying capacity of less fatty cheese, the butter-fat available for sale at butter prices would hardly cover the loss attendant upon the reduced output of cheese and the increased costs of manufacture. Finally it must be borne in mind that the profit arising from part-skimming will always depend upon the relative prices of butter and cheese and of ‘‘standardised” and “full cream” cheese. Assuming practically identical prices for the two classes, it is improbable that cheese would ever be so high and butter so low as to extinguish the profit completely. Reverting now to the question of yield, we may take it as established that the separation of milk, whether for starter or other purposes, must lower the yield of cheese per pound of butter-fat calculated in terms of the total butter-fat received into the factory. The process could, of course, be made to show on the balance-sheet, an increased yield, provided that the starter milk and all other milk skimmed, were credited to the account of “milk used for butter-mak-ing” and not to the account of “milk used for cheese-making.” The calculation of yield would then be made in terms of the butter-fat actually used in the cheese vats, and, under these circumstances, the published yield per lb of butter-fat would be falsely represented as higher than that actually obtained for all the butter-fat credited to cheese suppliers.

FACTORS INFLUENCING FALSE OR THEORETICAL CALCULATION OF YIELD.

Up to the present we have been considering factors which have a true influence upon the yield of cheese actually obtained, and which, by varying the weight of cheese manufactured, have a direct bearing upon the efficiency of the factory and the final prosperity of its suppliers. We shall now proceed to discuss the factors which have a false oi theoretical influence upon yield; which have no effect upon the weight of cheese actually manufactured, but which affect the yield merely as stated on paper, leaving the efficiency of the factory and the final prosperity of the suppliers exactly as before. In the first place, it must be borne in mind that the figure expressing the yield of cheese per pound of butter-fat is merely the result of a simple division sum, obtained by dividing the weight of cheese manufactured by the weight of butter-fat received. The quotient of any division sum can be increased either by increasing the dividend or by reducing the divisor. Applied to our own particular problem, this means that the published yield of a factory can be increased (on paper) with equal ease eithei by securing an increase in the cheese manufactured, or by effecting a reduction in the butter-fat credited. As an example of this false raising of yield, let us consider a normal milk having a true’fat test of 4 per cent. According to the standards already quoted, this milk will ’give’a true yield of 2.61 pounds of cheese per pound of butterfat. One thousand pounds of such milk would therefore contain 401bs of butterfat and would yield 104.41bs of cheese (i.e. 40 x 2.61). Suppose, however, that the true fat test of 4.0 per cent, were declared 3.9 per cent, and that only 39 pounds of butter-fat were credited instead of the 40 really received. . The same weight of cheese would still he made, and the yield per pound of but-ter-fat claimed would now be 104.4 divided by 39, i.e., practically 2.68. The yield figure has thus been raised ap<. parently by 0.07 pounds of cheese per pound of butter-fat without there being any more cheese available for sale, and thus without any influence upon the real prosperity of the suppliers. The standards already quoted have been formulated upon the basis of true butter-fat tests, all tests having been carried out. according to the official analytical method. The Babcock test, however, nearly always falls short of the analytical method by approximately 0.1 per cent., this tendency being accentuated by the recognised practice of giving “the factory” the benefit of all fractions in weight of milk and all fractions in test of butter-fat over the even decimal points. Hence it is correct to state that the butter-fat test credited at factories under present regulation metllO S is always, at least, 0.1 per cent, below the true analytical test, and this statement has been proved in innumerable instances. This has the effect of raising the apparent yields of cheese per poun of butter-fat in just the same way as in the example already quoted, and thus makes the standards appear as under:

It need hardly be repeated that these “apparent yields” are not actually obtained. They would, however, still be useful for comparing tlie working of one factory with another, provided that we could be certain that weighing, sampling and testing methods were universally identical. Unfortunately, there can be no guarantee that this is so.

Some managers may have their own particular methods of reading the Babcock bottle, giving less than the full official reading to the top of the extremities of the fat column. While this may inflict no great injustice upon the majority of a certain group of suppliers, treating them all alike, it nevertheless cannot be looked upon as a purely domestic affair and one which is solely the business of that particular company. The practice leads to errors in the calculation of the factory’s yield figure, which increase in proportion to the amount the test is “read down,” as shown in the following example: — Consider a milk with a true fat test of 4.2 per cent, and a true yielding capacity of 2.551bs of cheese per pound of butter-fat (raw milk cheese). By a calculation of the type already demonstrated, this true yield of 2.55 would appear as shown below if the fat test were “read down” to the figures indicated: — Fat test Apparent yield figure per read as lb of butter-fat. (True yield).

Hence, for every “point” that the average test is “read down,” the factory yield figure is apparently raised about 0:06 to 0.071bs of cheese per pound of butter-fut. It will be at once obvious that, where the possibility of this practice exists, the yield-figure may not give the slightest indication of the ability of the cheese-maker or of the efficiency of the factory. In fact, by a judicious application of the “reading down” principle, a measure of inefficiency could be completely masked by the publishing of a satisfactory or even high yield figure.

Such a state of affairs is not entirely the private affair of the company concerned. Yield figures now command widespread attention. They set standards which other factories, in self-de-fence, cannot ignore. Suppliers in certain areas will desert one company in order to supply another near at hand which has published a better yield figure in its balance sheet, and consequently paid a higher price per lb butter-fat. In order to regain its supply, the deserted company js compelled by hook or by crook to go one better. Directors have, no doubt unconsciously, egged on the contest by requiring candidates for managerial positions to produce evidence of yield, and have shown preference for the man whose figure was the highest. Under the combined stress of all these circumstances, it is not remarkable that published yields per pound of butter-fat have been steadily on the up grade for years in New Zealand, despite the fact that there has been a regular and real advancement in butter-fat tests, the true effect of which is a reduction in yield. Unless common sense and better knowledge prevail, it is difficult to forecast where the competition will end.

COMBINED EFFECT OF PART SKIMMING AND OF “READING DOWN” THE BUTTER-FAT TEST.

Before leaving this section of the discussion, we must consider another aspect of the artificial variation of the yield of cheese per pound of butter-fat, which will prove that the foregoing statements have some basis in actual fact. Additional evidence may be brought forward to show that, under some circumstances, not only may a high yield figure mean nothing at all, but also a normal yield figure need not necessarily be accepted as evidence that tests have been fully credited and cheese-making carried out according to orthodox principles. We have already discussed the effect of part-skimming, and have shown that this practice lowers the yield of cheese per pound of butter-fat. Furthermore, it will not be disputed that, in the past, all factories have, to a greater or a lesser degree, carried on a system of fat reduction which liar amounted in effect to part-skimming, whether by the legitimate process of starter skimming, setting aside of staff requirements in milk, or by the actual skimming of milk for use in the vats during the late autumn. These practices have all had their effect in lowering yield, and, if we are to believe some of the allegations made, some factories must have had a considerable leeway in reduced yield to make up if they have used sweet skim-milk in their vats even to a fraction of the extent alleged. Some explanation must therefore be found if we can. discover no case where the published yield is substantially below normal. The following example will show at least one method by which this can be done. Consider lOOOlbs of milk having a true test of 4.3 per cent, and a true yield of 2.521bs of cheese per lb of butter-fat. Suppose that this were skimmed to a test of 4.1 per cent, by the removal of 21bs of butter-fat. The true yield would now be 2.581bs of cheese per lb of but-ter-fat for the 411bs remaining, and thus the milk would manufacture 105.8 lbs of cheese (2.58 x 41). Calculated in terms of the original 431bs of butter-fat, however, the yield per lb is now 105.8 divided by 43, i.e., 2.4Glbs of cheese per lb of butter-fat. Suppose now, that in addition to the above skimming, the original test had been announced as 4.2 per cent, instead of 4.3 per cent., a contingency which is quite possible under the present system of Babcock testing. The butter-fat in the original milk would now be considered 4’2lbs and the apparent yield per pound of butrer-fat would now be 105.8 divided by 42 and equals 2.52, i.e., the same as the true yield of the original milk. If the test were “read down” to 4.1 per cent., the yield would similarly become 2.c3, and if “read down” to 4.0 per cent, the yield would (on paper) rise to 2.C5. From this example we may deduce the approximate rule that every 5 per cent, of part skimming will mask the effect of “reading down” the test 0.1 per cent, of butter-fat. Conversely, by extension of the rule, we see. that, by “reading down” the average test 0.2 per cent., about 10 per cent, of part-skim-ming could be done without the slightest lowering effect upon yield. The argument presented in this, section therefore shows that if, coincident with the practice of part-skimming, a high yield per pound of butter-fat is claimed, this can only be attained by an appreciable reduction in the test. Similarly it shows that, even if the yield finally presented on the balancesheet is about normal, there ma.v always be the suggestion that this Is tile result of a judicious combination of part-skimming and test-reduction. SUMMARY OF NATURAL AND ARTIFICIAL FACTORS INFLUENCING YIELD.

We ma ' now enquire whether any process of calculation will serve to disclose whether a yield figure has been subject to any artificial variation. To a very limited extent this may be answered in the affirmative, but compensating factors may be easily introduced, leaving everything apparently normal J . ■ ■ ■

If rich milk were being handled, one would naturally expect a low yield of cheese per pound of butter-fat, and also a low figure (approximately from 8 to 9) for the number of pounds of milk required to make one pound of cheese. If, however, the yield figure were artificially raised by “reading down” the test, this would not influence either the weight of milk received or the weight of cheese made. In consequence, the practice of “reading down” the test would be disclosed by a high yield figure and a low milk-cheese figure.

Another consequence of the “reading down” process would be the raising of the figure for manufacturing costs per lb butter-fat. While the scope of the factory and the weight of materials handled would remain constant, the but-ter-fat would be falsely reduced, thus raising the figure for “manufacturing and all costs to f.0.b.” per lb of butterfat. A scrutiny of this figure might therefore possess a limited value, but it would 1c .d to no certain knowledge because factories differ widely in their positions and natural advantages, thus making a standard for costs impossible. Moreover no real uniformity of practice has yet become universal in the compiling of dairy company accounts. Part’-skimming has been shown to lower the actual weight of cheese that could be made from a given quantity of milk. In consequence, it would lower the yield figure and increase the number of pounds of milk required 'to make one pound of cheese. Simple part-skimming would therefore be disclosed by a low yield and a normal or high milk-cheese figure, the actual range of the quantities being governed by the fat test of the original milk. Up to the present, no system of calculation has been devised which will effectively demonstrate part-skimming and “reading down” of test when both are carried on simultaneously and equally. Each has been shown to have a compensating effect upon the yield per pound of butter-fat, and similarly while “reading down” the test would leave the milk-cheese figure in the low region proper to fairly rich milk, part-skimming of such milk would raise the figure into the normal region of 9to 10. A scrutiny of costs would also give negative results because, for reasons already given, the raising of the “costs to f.0.b.” per lb of butter-fat which would accompany part-skimming and test reduction, either separately or conjointly could not be detected with certainty. When one considers that, in addition to these artificial methods whereby the yield figure can be raised, lowered or adjusted ip normal, there are numerous natural* factors (generally of an unknown quantity) pertaining to inilk composition and factory efficiency which have a true influence upon yield, it is not surprising if one comes to the conclusion that the yield figure may mean nothing at all, and that it certainly cannot be taken as an index of the efficiency of the factory or the cheese-maker unless all the influences which bear upon it are taken into account as indicated. [The original report contains at this point a table presenting in condensed form the logical interpretation which, it is claimed, can be legitimately placed upor the simultaneous occurrence of various combinations of high, low or normal yield and milk-cheese figures.]

INFLUENCE OF YIELD UPON PAYOUT PER POUND OF BUTTER-FAT AND UPON FARM RETURNS AND FINAL PROSPERITY OF THE SUPPLIERS.

Enough has now been said upon the subject of yieid to make it plain that the casual observer is not in a position to know the number, nature and magnitude of the various raising and lowering forces which have eventually fixed a yield at a certain figure. It is therefore fallacious to suppose that the highest yield is necessarily the best under the circumstances obtaining, and that consequently the highest pay-out per pound of butter-fat is the one which means the greatest net return to the farmer. An example will make this clear. Suppose that a factory had for a season a true average test of 4.2 per cent., a true yield of 2.551bs of cheese per pound of butter-fat and made a pay-out of Is 6d per pound of butter-fat. A farmer having produced IffiOOOlbs of but-ter-fat for the season and having been credited with his full weight of butterfat would receive a cheque for £l5O.

Suppose, on the other hand, that the test had been reduced by some process and its average made to appear 4.0 per cent. The company would now be able to announce its yield at 2.68, and in consequence, its pay-out per pound of butter-fat would rise to Is 6.9 d, very nearly a penny increase. A 10,0001 b butter-fat supplier, however, would share in the “reading down” process, and his butter-fat for the season would be reduced to 95251b5. His cheque would now be 9525 x Is 6.9 d, i.e., £750 exactly as before. This example shows that the supplier has not been robbed of any of his income, and that, although he received a higher price per pound butterfat, it was paid on fewer pounds of butter-fat, thus making his total income exactly the same by the two methods of calculation. Under the co-operative system, the suppliers continue to get the full proceeds of sale, less the costs of manufacture. If, therefore, their tests become reduced in a proportionate manner by any process, the net result is. a similar proportionate rise in the price paid per pound of butter-fat, lca\ ing the gross return exactly as before. Expressed in its simplest form, it is manifest that four pounds of butter-fat at Is 6d are equally as valuable to the farmer as three pounds of fat at 2s. The higher price, however, creates the impression of superior efficiency and great desirability, and may attract suppliers who are unaware that it may not necessarily put a penny more in their pockets. It will hardly be denied that the final measure of the prosperity of the cheese supplier is the return which he secures

per acre. This paper is an endeavour ta prove that the return per acre is not necessarily dependent upon the yield or the price paid per pound -of butter-fat by the cheese factory supplied, and that full weight must be given to a larg® number of considerations before placing an interpretation upon factory cheesemaking statistics.

-Sp ci Q Factory A high yield Factory B low yield CJL<y 'H. s ci m o’ ”2 Er o S3 in o a S 3 a> <p ’o a w 6 ft Er Nov., 1927 10 34.1 36.6 8 33.1 35.3 Dec., 1927 6 34.0 35.0 Jan., 1928 10 35.6 35.7 6 34.8 35.6 Feb., 1928 9 35.2 36.2 Mar., 1928 10 36.5 34.7 14 37.3 34.1 Mar., 1928 10 34.8 35.6 36.8 33.8

tfe. hb. kS>. 1U £ g g £ “ “Fat test. IO tO tO tO tO oi 5 to S! k Yield at factory to to to to to to . . . , / cioocjh weight cc h- -m © o< (p aw m ilk). to to to to to b*- b* bl Ol bl -4 C© )_ 4X Yield at factory to to to to to to . . . ,2 b c> Q packing weight o w c» o to -5 (Pasteurised, milk)

tO •M 4br* tO H4-- oo eo to to to True fat test. vield c a a" 2o’ 2 r/j fe IO bi to to 0'1 to oo co co w w Apparent b< Jp. co io h—» c cc do ° fat test. P to to to to to to to to to f® J® Apparent yieid o' o bi bi 00 bi ►—» •-4 b— 1 **• oo w raw milk. tO to to K> bO to to to to i® Apparent yield □o' to bi Ci bi —I c ■ 3 ' o o . o ot Pasteurised milk ‘

4.2 2.55 4.1 2.G1 4.0 2.68 3.9 2.75