ISSN:0003-2654    

DOI:10.1039/AN95277FX037

出版商:RSC    

年代:1952

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95277BX039

出版商:RSC    

年代:1952

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95277FP113

出版商:RSC    

年代:1952

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN95277BP121

出版商:RSC    

年代:1952

数据来源: RSC

摘要:

OCTOBER, 1952 THE ANALYST Vol. 77, No, 919 EDITORIAL INTERNATIONAL CONGRESS, OXFORD, 1952 THE First International Congress on Analytical Chemistry to meet under the patronage of the International Union of Pure and Applied Chemistry was held at Oxford from September 4th to 9th, 1952. The President was Sir Robert Robinson, OM., D.Sc., F.R.S., and the Vice-president was Sir Wallace Akers, C.B.E., F.R.I.C., F.R.S. The number of registered members of the Congress was 689, from 26 different countries. The Congress owed its inception to the expression, at the gathering of Analytical Chemists at Utrecht in 1948, of the hope that the next such gathering should take place in England. This desire was brought to the notice of the Society of Public Analysts and Other Analytical Chemists; and shortly afterwards, Dr.G. Roche Lynch invited a number of distinguished chemists to a meeting, which was held in the rooms of the Royal Society under the Chair- manship of Sir Robert Robinson. Committees were appointed to take active steps to further the matter and the suggestion that had emanated from Utrecht was developed towards maturity. Mr. R. C. Chirnside, F.R.I.C., became the Honorary Secretary of the Congress, and Mr. George Taylor, O.B.E., F.R.I.C., was made Chairman of the Executive Committee. The Committee turned to our Society for some preliminary assistance, and the Society pledged the help of The Analyst in publishing the papers and discussions. A brief glance a t the committees listed in the Congress Handbook shows the extent to which leading members of the Society have given freely of their time.The Congress was recognised from its inception by the International Union of Pure and Applied Chemistry and was, in effect, the first Congress of the Analytical Section of the Union. The Congress proved immensely successful and we hope it will form the pattern for future Congresses; one of its most notable features is the rapidity with which the proceedings are being published. The papers had to be prepared before- hand so that they could be set in type before the event, and publication had to be carried out by already-organised machinery. That this has been accomplished is in no small measure due to the diversion of the efforts of the Society’s permanent staff and the facilities afforded by the printers of The Analyst. h d although, by pre-printing the scientific papers, work on the publication was effectively started on May 8th, it is by no means finished as this issue appears, for the Congress Lectures could not be put into type before the meetings at Oxford, and the discussions, by their very nature, were available only after the meetings.The willing help of the contributors to the discussions and of the authors in preparing their replies has greatly assisted the task that we have faced. The November issue will contain a report on the (hgress, the three main Congress Lectures, and the first part of the Scientific Papers and tlie discussions on them; the December issue will contain the balance of the papers and discussions. All other matter has had to be excluded from these two issues. It is hoped that the bound Proceedings o€ the Congress, which will include the papers appearing in the Xoveniber and December issues together with their own index, will be available early in 1953. This has involved two factors. The fruition of these plans will he apparent in the next two issues of The A d y s t . 497

ISSN:0003-2654    

DOI:10.1039/AN9527700497

出版商:RSC    

年代:1952

数据来源: RSC

摘要:

498 PROCEEDINGS [Vol. 77 PROCEEDINGS OF THE SOCIE’T‘Y OF PUBLIC ANALYSTS AND OTHER ANALYTICAL CHEMISTS NORTH OF ENGLAND SECTION AN Ordinary Meeting of the Section was held ai: the Engineers’ Club, Albert Square, Man- Chester, on Saturday, April 26th, 1952, at 2 p.m.. The Vice-chairman, Mr. T. W. Lovett, F.:R.I.C., presided over an attendance of 58. The following paper was presented and discussed: “Some New Methods in Analytical Chemistry,” by R. Belcher, R.Sc., F.R.I.C., F. 1nst.F. THE Fifteenth Summer Meeting of the Section was held at the Imperial Hotel, Llanduclno, from Friday, June 13th, to Monday, June 16th, 1952. The Chairman, Mr. A. A. D. Comrie, B.Sc., F.R.I.C., presided over an attendance of 40. On the morning of Saturday, June 14th, a paper was read by Dr. J. Haslam, F.R.I.C., on “The Analyst in the Plastics Industry” and a number of plastic products now in common use were exhibited.A tour of the district was :made by motor coach on Sunday afternoon. BIOLOGICAL METHODS GROUP THE Summer Meeting of the Group took the form of a visit to the premises of Boots Pure Drug Co. Ltd. In the afternoon of June 12th, 1952, visits were made to the Pharmacology, Bacteriology and Bio-assay Divisions at West Bridgford. The following morning the party visited the Beeston factory, and after lunch, which was kindly provided by Messrs. Boots, was shown the Veterinary Science Division. Finally, a scientific session was held, at which the following papers were presented: “The Assay of Adrenaline in Adrenaline Cream,’’ by G. F. Somers; “The Assay of Pertussis Vaccines for Antigenicity and Toxicity,” by P. W. Muggleton. Shortage of time prevented a discussion of the second paper and a demonstration of the “Evaluation of Histamine-Release by Skin Test:; in Volunteers, Introducing a Rapid Method for Measuring Small Irregular Areas,” by H. 0. J. Collier and Barbara Macauley.

ISSN:0003-2654    

DOI:10.1039/AN9527700498

出版商:RSC    

年代:1952

数据来源: RSC

摘要:

Oct., 19521 DAVIS 499 The Chemical Between Composition of 1900 and 1950 Milk BY J. G. DAVIS (Presented at the meeting of the Society on Wednesday, November 7th, 1951) It has been suggested, during the past few years, that the chemical quality of milk in this country has deteriorated over the last twenty or thirty years. The difficulties encountered in an attempt to answer this question are discussed in some detail. One of the most important requirements is the systematic testing of samples selected at random all over the country by the same analytical methods, over the period under consideration. There is insufficient information to allow any dogmatic assertion about the trends in milk quality. In the present paper, the results of over half-a-million samples tested in various parts of England during the period 1900 to 1950, have been tabulated and examined.It appears that there has been a slight rise in fat content, with a peak at 1926 to 1930, and subsequently a slight fall. The solids-not-fat values appear to have risen steadily from 1900 to 1921-1925 and since then to have fallen appreciably. Further evidence is produced to suggest that there has been a similar fall in the quality of milk in Wales, whereas the fat content of milk in Scotland has increased. Methods for maintaining or improving the quality of milk are discussed, and it is suggested that the logical solution is a universal method of payment by quality. The total solids content is suggested as being the most suitable from chemical aspects and also from the points of view of justness, simplicity, and methods of utilisation.The evidence available from other countries suggests that where milk is bought on a quality basis, the fat and solids- not-fat contents are at least maintained, if not improved, and where payment is independent of quality, the chemical composition tends to deteriorate. THE trends of composition of our national milk supply obviously depend on the methods used to obtain analytical data. No conclusion can be drawn from any figures unless the methods of sampling are above reproach and the analytical methods used have been con- sistent throughout the period under consideration. It is not necessary that methods be absolutely accurate in order to decide whether the composition has altered. SAMPLING- It might appear that if all herd milks in the country had been tested regularly over a period, it would be possible to answer this question.A little thought will show, however, that a number of factors can affect the result even if the sampling is, from the scientific point of view, constant. For example, let us suppose that at the beginning of the period all herds in the country were of equal size, one half giving richer milk having average values for fat and solids-not-fat of 4.0 per cent. and 9.0 per cent., and the other half giving a poorer milk with averages of 3.5 per cent. and 8.5 per cent. of fat and solids-not-fat. If the herds giving the poorer milk doubled in size, the final averages of the country’s milk would then be not 3-75 per cent. and 8.75 per cent. but 3.67 per cent.and 8.67 per cent., i e . , a difference in both figures of nearly 0.1 per cent. This example will show that any method used for assessing the chemical composition can be criticised from some point of view. Averages taken from values for herd supplies will not necessarily be the same as those found by random sampling of tanker or consumer milk. From the point of view of analytical control, milk can be sampled in three ways: (1) as produced by the cow, (2) as sold from the farm and (3) as sold to the public. Obviously, to assess trends in composition accurately, milk should be tested as produced by the cows, but analytical figures for this type of milk are quite inadequate for the present purpose. Studies made by dairy scientists on trends in the composition of milk are usually based upon milk as sold from the farms or as bought by collecting depots, while statements500 DAVIS : THE CHEMICAL COMPOSITION OF [Vol.77 made by Medical Officers of Health and Public Analysts are usually based upon the testing of consumer samples. We have no precise knowledge about the relationship between these three types of milk, but we know that any differences, if they occur, will always be in the direction of poorer milk as we pass from (1) to (2) and from (2) to (3). HUMAN FACTORS AFFECTING COMPOSITIONAL DArA- The handling of milk is always accompanied by the possibility of weakening of com- position, and this can occur in two ways: by the addition of water and by the abstraction of fat. I t might generally be thought that adulteration is, in practice, the major factor, but some consideration of the matter will show that abstraction of fat may be more important from the point of view of general composition. For example, all dairy farmers use some of their milk to supply their households and staff, and if milk for this purpose is taken from the top of an unplunged churn of evening milk, the milk removed will be much higher in fat than the remainder.A simple calculation will show that, for an average farmer producing daily 20 gallons of milk containing 3.75 per cent. of fat, if four pints of milk richer in fat are removed every day, the fat content of the remainder will be reduced to 3.5 per cent. if the milk removed contains 10 per cent. of fat, arid to 3.25 per cent. of fat if the milk removed contains 20 per cent. of fat.It is emphasised that we have no precise knowledge of the magnitude of this operation on the fat content of the country’s milk supply, but estimates that I have received from dairy farmers indicate that the extent of this practice is about 90 per cent. It is a very old- established practice and this figure is probably about right. Precise information is only available in isolated instances, but some interesting data illustrating this point are shown in Table I. This table gives average values for large numbers TABLE I EFFECT OF PLACE OF SAMPLING ON FAT CONTENT A B C D Sampled a t farm or at creameries Sampled at schools Retail samples Sampled in shclys Year - -7 +--7 - Fat S.N.F Fat S.N.F. Fat S.N.F. Fat S.N.F.1948 4.07 8.83 3.65 8.80 3.74 8.86 3-68 8.78 1949 4.06 8.83 3-73 8.75 3.67 8.89 3-65 8-84 1950 4-16 8.89 3.77 8.89 3.77 8.92 3.72 8.82 of samples taken by one authority under different conditions, group A samples being taken at the farm and groups B, C and D being taken at the point of distribution or consumption. It will be observed that, whilst the solids-not-fa.t values are remarkably even, the fat values for the samples taken at the farm are much higher than those taken during distribution. It would not be justifiable, of course, to suggest that these figures are true for the whole country, but they illustrate the point that the place of taking the sample may have an appreciable effect on the fat but not on the solids-not-fat values. Further, farms are not the only places at which abstraction of fat may take place. Public Analysts who worked in the earlier days of the dairy industry will recall how remarkably consistent some retailers were in selling milk containing just over 3 per cent.of fat. Although not manufacturers and not equipped with a laboratory, such dairymen frequently possessed a Gerber outfit. Abstraction of fat by removing a portion of the milk that is richer in cream will have only a negligible effect on the solids-not-fat value, while adulteration of milk by water, which is not always intentional, will affect fat and solids-not-fat equally, considered as a percentage of the material present, but in terms of percentage of the milk will have rather different effects. For example, the addition of 1 per cent.of water to a milk will lower 3.75 per cent. of fat to 3.71 per cent. and 8-75 per cent. of solids-not-fat t o 8.66 per cent., while a 5 per cent. adulteration will lower these values to 3.56 and 8.31 per cent. From the point of view of the absohte change in percentage of the constituents, therefore, the addition of water will affect solids-not-fat values more than fat values.act., 19521 MILK BETWEEN 1900 AND 1950 501 FACTORS AFFECTING THE COMPOSITION OF MILK It is well known that a number of factors can affect the composition of milk from day to day and season to season. These, however, level out when average values for years, or groups of years, are considered, so that such factors need not be discussed in the present paper. Further, it is known that the milk from one cow, or even from one herd, can vary in composition from day to day, especially in respect of fat content, but these individual variations, and variations for which no precise cause can be ascribed, do not come into the consideration of yearly values.There are, however, some very important factors, not only physiological but economic and industrial, that play their part in affecting the composition of milk over a period of years, and these will now be considered. CHANGES IN THE RELATIVE NUMBERS OF DIFFERENT BREEDS- It is well known that breed is a most important factor, and typical figures are shown in Tables I1 and 111. No accurate figures are available for the total numbers of cows of each breed in Great Britain, but Table IV, which gives the numbers of licensed bulls of each breed, allows us to make an estimate of the relative numbers of cows of each breed.TABLE I1 REPORTED AVERAGE FAT AND S.N.F. VALUES FOR BREEDS .. 1880 1885 . . Penn. State Barham Fair * .. 3.89 .. 5.57 .. 3-58 2.97 8.83 .. 5-50 4.8 9.29 .. 3.91 4.26 9.34 . I 3.79 8-91 1899 Richmond Date Source 1899 Lewin (Bell) Richmond Ayrshire .. Devon . . .. Friesian .. Guernsey . . Shorthorn . . South Devon . . Jersey . . . . Date Source 3.68 9.02 4.24 9.22 3-51 8.61 5.02 9.46 4.78 9.56 3.65 8.80 3.75 8-65 5.16 9.3 5.43 9-22 3.93 8.85 5-66 9-23 4.03 8.85 .. . . .. .. .. .. .. .. . . 1899 State Exp. Sta.* Collier* New Jersey 3.57 9.45 3.68 9.02 4-15 9.59 3-46 8.96 3-51 8.61 5-12 9.47 5.02 5-61 9.80 4.78 9.56 3.65 8.80 1914 Lythgoe* 1909-21 B.D.F.A.? Milking Trials 9.05 9.13 8.74 9.26 9.20 9.08 9.16 1929 Overman et al., Illinois* 4.14 8-94 1920-22 Essex County Milk Records A yrshire ..Devon . . .. .Friesian .. Guernsey .. Jersey . . .. Shorthorn . . South Devon . . Date Source 4-01 8-63 3.41 8.28 5.23 9.37 5.65 9.10 3.36 3.70 . . .. 1925 1926 . . Tochcr Bibby 1927 Drakeley 1936 Davies Ayrshire .. Devon .. .. Friesian .. Guernsey . . Jersey . . .. Shorthorn . . South Devon . . Date Source . . 4-09 3-9 .. 4.5 . I 3.63 3-45 . . 5.16 5.31 .. 5.43 5.56 .. 3-91 4.14 3.97 9.09 3-85 9.05 3-67 8.78 4.88 9.29 5.18 9-30 3.78 9.00 4.02 9-09 1948-49 3.55 8-97 5-19 9.68 5-18 9.51 3.45 8.48 4-98 9-67 5.14 9-59 3.63 8.94 .. . . 1948-49 1948-49 . . West Region South-Eastern 1949 3i.I.R.D.Z 3.72 8.78 3.49 8.59 4.55 9.01 3.56 8-71 1950 Bartlett National Region Milk Records .. 3.88 3.76 3-81 .. 4.18 . . 3.51 3.41 3.45 .. 4.65 4.52 4.56 .. 5.1 1 5.00 5.05 .. 3.65 3.52 .. 4.26 Ayrshire .. Devon . . .. Friesian .. Guernsey . . Jersey . . .. Shorthorn . . South Devon . . 3.60 9.05 3-88 9-13 3.47 8-74 4.53 9.26 4.86 9.20 3.63 9.08 3-81 9.16 * American figures, t British Dairy Farmers' Association. 1 National Institute for Research in Dairying.502 DAVIS THE CHEMICAL COMPOSITION OF [Vol. 77 It may be remarked in passing that the figures taken for the over-all breed average composition are the figures for registered or otlnenvise selected herds. Nevertheless it is reasonable to assume that these figures, although selected, apply generally to all the cows of each breed in the country.Jersey and Guernsey are, beyond question, the richest milks, and Devon, South Devon and Ayrshire are usually richer than the average, Shorthorn milk approximates closely to the average milk supply of the country, whilst Friesiae milk is generally poorer in composition. Other things being equal, therefore, any trends in the relative numbers of cows of different breeds might be expected to affect the averagn TABLE ]:TI YIELD AXD FAT PERCENTAGE OF MILKS OF VARIOUS BREEDS ED WARDS^) Breed or breed type hyrshire . . .. . . Friesian . . .. .. Guernsey . . .. .. Jersey . . .. .. .. Red Poll . . .. .. Shorthorn . . * . .. Breed or breed type Ayrshire . . .. Friesian .. . . Guernsey . . .. Jersey . . .. .. Red Poll . . .. Shorthorn . . .. TOTAL BULLS Breed Ayrshire .. .. Guernsey . . . . Jersey . . .. . . Red Poll . . .. Shorthorn . . .. Aberdeen Angus . . Hereford . . .. Others .. .. Total . . .. Friesian . . . . .. .. . . .. .. .. Number of cows considered 7- 1934-35 1948-49 . . 3278 43,112 . . 13,438 70,657 .. 6224 13,183 .. 3797 8458 .. 3742 6147 . . 41,726 73,456 Number of cows and heifers (tested) 7- 1934-35 1948-49 . . 906 50,673 . . 2908 71,779 . . 2639 18,456 . . 1296 13,324 .. 650 6333 .. 2822 36,260 TABLE I V Yield of milk, lb. & 1934-35 1948-49 7514 8088 9291 9141 6891 7473 6635 7051 7589 7476 7318 7365 Fat, % -7 1034-35 1948-49 3.71 3.81 3-26 3.45 4.63 4.56 3.86 5.05 3-65 3.60 3.56 3.58 LICENSED IN ENGLAND AXD WALES (1934 TO 1951)* 1934 1941 1945 1950 . . .. 192 750 2684 3115 . . .. 1210 6125 8220 9308 .... 969 1037 1219 1582 .. .. 328 369 548 944 . . . . 417 377 467 425 . . . . 13,908 21,023 14,593 7902 . . . . 635 512 458 439 .. .. 1806 2191 1740 1784 .. .. 3142 4172 3260 2043 36,556 33,189 27,542 .. . . 22,607 * Figures from the Ministry of Agriculture and Fishcries. composition of the country’s milk. It is known that the Shorthorn has been the predominant breed in this country from the time when analytical data for milks first became available until fairly recently; the numbers of cows of other breeds have been relatively small. How- ever, during the last 15 or so years there has been a significant change in the numbers of Shorthorn and Friesian cows and bulls in this country. Whereas 15 years ago the number of Friesians was relatively small, to-day there are roughly as many Friesian cows as there are Shorthorn.This fact does not, of course, ;brow that the quality of the country’s milk has necessarily fallen, as it is conceivable that the average values for Shorthorn milk have fallen over this period and the average values for Friesian milk have risen. However, there is a strong presumption that the increase in the proportion of the cows giving milk of poorer quality will result in a lowering of the compositj.ona1 analysis for the milk supply as a whole. Provan and Jenkins2 point out that the fat values for breeds, other than Channel Island, have remained stationary, the Channel Island value having fallen by 0-5 per cent. The figures for solids-not-fat have fallen for all breeds except the Friesian, which has increasedOct., 19521 MILK BETWEEN 1900 AND 1950 503 slightly.It is claimed that the difference in solids-not-fat between the Friesian and Shorthorn is now only about 0.1 per cent., and that the increase in numbers of Friesian cows cannot explain the observed fall in solids-not-fat values. THE EFFECT OF BREEDING OK SELECTION POLICY- Two factors in recent years have had the effect of making the country more milk conscious- (1) The foundation of the Milk Marketing Board in 1933, which organised the marketing of the country’s milk and has done much to publicise the value of milk. Advances in our knowledge of nutrition, which have made the public and all those concerned with health and feeding more conscious of the need for an adequate supply of milk for each individual.This was rounded off during the war when milk was given a high priority and special arrangements were made for ensuring that those most urgently needing milk got an adequate supply without any financial burden. Thus slightly over half of the country’s milk goes to the priority classes and the average consumption has risen from about 0.37 pints per day in 1938 to 0.75 pints per day in 1950. The new marketing arrangements have been accompanied by an appreciable increase in the price of milk paid to the farmer, with the result that milk production has become a much more profitable and less hazardous business than it was before 1933, when surplus milk in the flush period might be sold for as little as 6d. or 4d. a gallon. The value of milk as a marketable product has naturally resulted in farmers paying more attention to it as a business proposition, and there has been generally a marked increase in the efforts of farmers to produce the largest possible quantity of milk from any one farm.I t will be obvious that the greater the quantity of milk produced from a single cow the more economic will be production of the milk, as most of the overheads remain roughly the same. Considerable attention has been paid by farmers to producing milk from high-yielding cows, and this principle applies not only to Friesian, but to all breeds. This has resulted in a greater interest in Friesians and high-yielding Shorthorns, and the increased attention that is paid to production figures for cows and herds has been concentrated on production as measured in gallons.The farmer is paid for his milk on a quantity basis, and provided that it contains at least 3.0 per cent. of fat and 8.5 per cent. of solids-not-fat and, of course, is of satisfactory bacteriological quality, the milk must be accepted without question by any buyer. In fact, a large amount of milk containing less than 8.5 per cent. of solids-not-fat is now accepted, especially in late winter and early spring. From the business point of view an intelligent farmer would merely take precautions to ensure that his milk complied with the presumptive legal requirements and from that point concentrate on the production of the greatest possible quantity. Farmers who adopt a scientific and systematic breeding policy would naturally base the policy on the production of milk in terms of gallons rather than in terms of milk quality.It has been suggested by Kay3 that these two factors, the increase in the proportion of Friesian cattle and breeding for quantity without regard to quality, have been the two most important factors responsible for the alleged decline in the composition of milk in this country. Trends in yield by cows of different breeds are shown in Table I11 (p. 502). MASTITIS- There has been much conflict of opinion on the quantitative significance of mastitis in dairy herds as a cause of low compositional quality. About 1938, surveys indicated that roughly one-third of the dairy cows in milk suffered from mastitis in some form or other.*j5 Only about 5 per cent. of infected cows showed clinical symptoms and so would be recognised as infected by the farmers and milkers. Nevertheless, it was shown that sub-clinical infections could result in a lowering of the compositional quality of the milk, especially the solids-no t -f at .6 9 Considerable advances have been made recently in combating mastitis with penicillin and other antibiotics and drugs, and it is probable that today the average incidence of mastitis (of all types) is something betweeii 15 and 10 per cent.I t is reasonable to assume, therefore, that the mastitis effect, whatever its magnitude, is to-day only one-third to one-half of what it was about 1938.504 DAVIS : THE CHEMICAL COMPOSITION OF [Vol. 77 MACIIIXE MILKING- The fat content of milk is considerably affected by the efficiency of “stripping,” or removing the last amounts of milk from the udder, as this fraction is the richest in fat.Such “strippings” may contain 13 per cent. of fat and amount to one-fifth of the volume of an efficient milking. Machine milking may result in a less efficient removal of the last milk, especially if hand-stripping is not done properly or is omitted. Only fat values will be affected. DECREASE IN THE NUMBER OF PRODUCER - RETAILERS AND INCREASE I N WHOLESALERS- The last 50 years have seen a significant revolution in this country in the methods of selling milk. Whereas in early days milk was sold direct by the fa.rmer to the public-the farmer was running his own retail round-to-day the bulk of the milk is collected from the farm by the buyer’s lorry and taken to a collecting depot where all supplies are regularly tested.There was, therefore, in the early days of the dairy industry, little check on the quality of milk as sold to the public, apart from the occasional sampling by Food and Drugs authorities. Even then the “appeal to the cow” method still permitted a milk producer to sell milk below the legal presumptive standards if the authorities were satisfied that the milk was genuine. To-day, however, nearly all milk supplies going to a country creamery are tested regularly, usually once a month, in the buyer’s laboratory, and any deficiencies or near-deficiencies in fat and solids-not-fat are, or should be, immediately brought to the farmer’s notice. Quite apart from the reporting of low fat and solids-not-fat values, there is no doubt that the mere fact that a producer’s milk is tested regularly will have some influence in making the farmer pay more attention to the quality of his milk. Hence, the present system by which milk is bought will act in one direction by making the farmer interested in obtaining high yields without much consideration of quality, but the collecting and testing system under which the milk is marketed will always pull him up if his milk falls below a certain level.Thus, although the general average may have fallen, there are probably far fewer herd supplies of grossly poor quality. Public Analysts have commented on the fact that the passing of milk sales to the big distributors has usually resulted in a decrease in the number of deficient and adulterated samples.This point is illustrated by figures for a London borough (bulked milk) and a Home County (mostly producer-retailers) in 1948. Of 479 samples taken in London, not one was deficient in fat or contained added water (when tests were made). Of 326 samples taken in the county, 7 contained added water, 18 were deficient in fat and 4 were both adulterated and deficient in fat. Altogether, 3.4 per cent. were adulterated and 6.8 per cent. were deficient in fat. THE EXTENT OF ADULTERATION OF MILK- From the Public Analysts’ point of view adulteration can take place at any time between the milk leaving the cow and the inspector taking the sample. For the creamery testing milk as sold from the farm, adulteration is limited to any that takes place on the farm and is therefore likely to be less, but in both instances the extent is unknown.The freezing-point test did not come into general use until 1925 to 1935, and even then only a minimum adultera- tion of 3 per cent. could be detected. SEASON AND FEEDING- Reference is made on p. 505 and 520 to low values for solids-not-fat. The large mass of evidence on this topic is hopelessly confusing and contradictory. EARLY INFORMATION ON TH‘E COMPOSITION OF MILK Analyses of milk before 1900 must be treated with suspicion for more than one reason. Methods then in use are now known to be inaccurate, and the sampling and degree of adultera- tion were then probably greater sources of error than they are to-day. Wanklyn’s fat method, which was semi-officially adopted by the Society of Public Analysts in the 1870’s, gave very low fat values (as much as 0-5 per cent.too low) and therefore correspondingly high solids- not-fat values, although the total solids values were also possibly high at this time owing to failure to drive off all the moisture. One of the earliest tables recording breed values is that given by Barhams in 1885 for a 5-year average of British Dairy Farmers’ Association figures, reproduced in Table V. TheOct., 19521 MILK BETWEEN 1900 AND 1950 505 low fat and high solids-not-fat values for the milks of Crossbred and Dutch cattle are remark- able, but it is likely that these fat values are all low and the solids-not-fat values all high. The various editions of Richmond’s “Dairy Chemi~try”~ and InglelO give the average composition of milk as shown in Table VI.TABLE V BREED VALUES RECORDED BY BARHAM XK 1885 Breed Fat S.N.F. Guernsey . . .. .. . . 4.8 9-29 Shorthorn . . .. . . . . 3.79 8-91 Crossbred . . .. .. . . 3.15 9.16 Dutch (Friesian) . . . . . . 2.97 5-83 Jersey . . . . .. . . . . 4.26 9.34 As the Vieth and Richmond results were taken from samples received at a creamery, and Macdougal’s results were from commercial samples, the difference gives some indication of the extent of adulteration and abstraction of fat. On this straightforward assumption, the difference in fat figures indicates an over-all adulteration of about 2Q per cent. and the difference in solids-not-fat figures about 3$ per cent. This estimate should be regarded with caution, as the figures are not from the same herds or over the same years.It is significant, however, that the later the point of sampling, the lower are the average values. TABLE VI AVERAGE COMPOSITXOX OF MILK ACCORDING TO RICHMOSD9 AND INGLE” Richmond* A I -$ Ingle (Before 1899) (1897-1914) (1900-1920) (1 900) Fat . . . . . . . . . . 3.90 3.74 3.75 3.90 Solids-not-fat . . . . .. 9.00 8.91 8.91 9.00 Protein . . . . . . . . 340 3.40 3.40 3.40 Lactose . . . . . . . . 4-75 4.70 4.70 4.75 Ash . . . . . . .. . . 0.75 0.75 0.75 0.75 * Richmond’s values are based OD analyses of milk sent to the Aylesbury Dairy Co. Other data published about this time and quoted by InglelO are as follows- Authors No. of samples Fat Vieth and Richmond .. .. 200,000 3.9 Maxdougal . . .. .. . . 12,000 3.82 S.N.F. 9.0 8-68 Froni cow to depot there is evidence for a fall in fat and from depot to consumer there is evidence for a fall in both fat and solids-not-fat, which suggests that on the farm there was abstraction of fat and, by the dairyman at that time, there was an appreciable adulteration.The average values quoted by Pearmain and Moorll are 3.6 per cent. of fat and 8.8 per cent. of solids-not-fat, so that an average milk would have to be adulterated by 4 per cent. before detection by determination of solids-not-fat and by 31 per cent. before detection by determination of fat (2-75 per cent. taken as standard). Bell, in 1892, found average values of 4 per cent. of fat a d 8.9 per cent. of solids-not-fat, and Vieth, in 1892, found 4.1 per cent. of fat and 8-8 per cent. of solids-not-fat, according to Pearmain and Moorll (Zoc.cit., p. 14). TXE INCIDENCE OF MILK CONTAINING LESS THAI\: 8.5 PER CENT. OF SOLIDS-NOT-PAT During recent years, attention has been directed to the large proportion of herd supplies coming into creameries that are below the legal presumptive limit of 8-5 per cent. of solids- not-fat. Cranfield12 found that 16 per cent. of herd samples were below 8.5 per cent. of solids- not-fat in 1923 to 1925, and Burrl3 found that 13 per cent. in 1934 and 9 per cent. in 1936 fell below this standard. During and since the war this has been most noticeable in late winter and early spring. Rowland14 reported in 1944 that as many as 81 per cent. and 86 per cent. of samples at two creameries contained less than 8.5 per cent. of solids-not-fat at the times of lowest solids-not-fat .The solids-not-fat values rose abruptly in May.506 DAVIS: THE CHEMICAL COMPOSITION OF [Vol. 77 Although it was formerly commonly stated that feeding had no effect on solids-not-fat values, the shortage of concentrates during the war resulted in renewed investigation of this problem and it is now accepted that low protein, and especially low calorie intake, can result in low solids-not-fat values.15~16~~7 The effects of season and feeding cannot easily be separated, but recently Featherstone, Rickaby and Cavellls have shown that it is possible for a herd that is efficiently managed and well-fed to give milk low in solids-not-fat in late winter and early spring. Those outside the dairy industry may believe that this is a relatively recent observation, but this is far from the truth.It has been recognised for over 50 years that cows may secrete milk that has less than 8.5 per cent. of solids-not-fa.t, and evidence for this has been published by Richmond (quoted in the Wenlock Reportlg) and Tocher20, and a most useful summary of this particular problem was given by Hanley.21 This problem was fully recognised by the \Venlock Committee, and the Sale of Milk Regulations of 1901 recognised the so-called “appeal to the cow” test, which was designed to prevent the unjust conviction of a dairy farmer whose cows happened to produce milk low in solids-not-fat. The “appeal to the cow” test was, in fact, suggested many years earlier by a magistrate and was immediately put into use by some Public Analysts.With the state of scientific knowledge at the time, this was undoubtedly a most useful measure, because there was no test that could detect adulteration with any degree of accuracy. Probably, at that time, added water could only be detected when present to the extent of, say, 10 per cent. or more. The reasons for cows secreting milk low in solids-not-fat were but little understood in 1901. Presumably there was no trouble due to shortage of concentrates and other artificial foods at that time, as cows would have been fed almost entirely on grass and hay supplemented by other home-grown crops. However, sub-clinical mastitis, which is now recognised to be a considerable predisposing factor to low solids-not-fat in certain herds, was not recognised at this time, and this factor was ignored in papers on the subject until the 1930’s.THE FIXIKG OF PRESUMPTIVE: STANDARDS IN 1901 It is of interest to note that although there was considerable argument about the fat standard in the Wenlock Committee deliberations (11 witnesses recommended 2.75 to 2.8 per cent., 21 a value of 3.0 per cent. and 3 a limit of 3.25 per cent. of fat), the standard of 8.5 per cent. for solids-not-fat appears to have been unquestioned. This had been in use by the Society of Public Analysts for many years, although Richmond had recommended a limit of 8.25 per cent. for July to September. It is therefore clear that the present trouble with late winter and early spring low solids-not-fat wats unknown at that time. The variety of recommendations about the fat standard may have been partly due to the errors of the methods previously used (e.g., the Wanklyn method is now known to give results about 0.5 per cent.low compared with the Adams coil method, but this itself may give high results). It is also obvious that a method giving a low fat result would ;$so facto give a high solids-not-fat result, because the latter value was at that time calculated by the difference between fat and gravimetric total solids test<. Richmond made a statistical analysis of 76,058 fat tests taken over 6 years, but it is important to realise that this period of 6 years (1894 to 1899) is not necessarily an adequate sample in respect of yearly variations and, further, Richmond had no knowledge of the extent of adulteration of these samples.Richmond’s values yielded an average of 3-84 per cent. and a mode of 3.75 per cent. and there were about 700 results of less than 3 per cent. Richmond’s monthly figures show a minimum “standard” €at of about 2.8 per cent. in May and June, 1898, and a minimum “standard” solids-not-fat of 8-3 per cent. in August, 1897, but he had no real knowledge of the extent of either the abstraction of fat or addition of water. His standards were obtained by statistical treatment of his test results, equivalent to excluding the lowest 23 per cent. of the values (Hanley21j. DIFFICULTIES IN COMPARIXG 1900 DATA WITH DATA TO-DAE- There are two major sources of error in comparing the average composition of milk in 1900 with that of to-day. At that time the adulteration of milk was notably more extensive than it is to-day because the inspection and testing of milk was less efficient, and the freezing- point test was then not in use.Further, there was about this time a change in analytical methods. Previously, milk was tested almost universally by the weighed total solids andOct., 19521 MILK BETWEEN 1900 AND 1950 507 a direct fat method. It is obvious that, since the solids-not-fat was calculated by difference, the use of any method giving a low fat percentage would automatically give a high solids- not-fat percentage. The Wenlock Committee, which issued its report in 1901, appears to have assumed a value of 8.9 per cent. for the solids-not-fat content of milk at that time, and I understand that this was based mainly upon Richmond’s evidence and Ingle’s report.10 I t is possible that this value was high because the generally accepted values for solids-not-fat at that time were high on account of low fat extraction. THE IMPORTANCE OF ANALYTICAL METHODS In general the milk supplies of our country are controlled by two very simple routine sorting tests-the Gerber test for fat and the lactometer or “Richmond” test for solids-not-fat.These have been in general use by milk buyers for nearly 50 years and are to-day commonly used by Public Analysts for the routine or first testing of milk samples, the more accurate methods being used only when these routine tests indicate a deficiency. The Gerber test for fat can be briefly dismissed by saying that, properly carried out in a calibrated apparatus, it is a very accurate method, the error being at the outside 0.1 per cent.of fat. Users of the test will probably consider that their results are accurate to 0.05 per cent., but there should be no opportunity for errors of any magnitude. The test for solids-not-fat is a little more complicated and requires some discussion. The Richmond method derives the solids-not-fat value from the known fat content and the specific gravity of the milk. G 6F 4 5 was derived by Richmond about 1895 from his analysis of about 240 samples of milk at different seasons of the year. His formula was derived by the method of best fit to accurate analytical data. Unfortunately there are a number of sources of error associated with this method. In the first place, no two milks are identical in composition; it follows therefore that a specific gravity method can only be worked out to give an average value and the error, other things being equal, will be the greater the more abnormal the composition of the milkq9 Further, apart from milks of abnormal composition, there are other factors that can affect specific gravity, particularly the degree of aeration of the milk and the condition of the fat.Aeration such as might be caused by allowing milk to fall into a tanker can result in a decrease in specific gravity equivalent to as much as 0-15 per cent. of solids-not-fat. If it is assumed that the aeration effect is nil and the milk is of average composition, there still remains the most important disturbing factor, that due to the slow solidification of the fat in milk when it is held for some hours at low temperatures.Although milk fat melts at about 32” C, solidification is still a slow process at temperatures much below 32” C. This effect, commonly called the Recknagel effect, can result in a difference in specific gravity equivalent to as much as 0.2 per cent. of solids-not-fat. It is a matter of considerable importance, therefore, to standardise the condition of the fat globules. It is not clear whether the Richmond results were obtained on milks in which the Kecknagel effect had not taken place or on milks in which it had been completed, and sub- sequent investigations of this problem still leave the matter in some confusion. Some years ago Boden and Campbell,22 in Scotland, found that refrigerated milk samples gave correct results for solids-not-fat by the lactometer method and they therefore assumed that the Richmond formula was intended to be used with milks in which the Recknagel effect had been completed. At Reading, however, R ~ w l a n d ~ ~ found that the Richmond lactometer method used on refrigerated milks (ie., with the Recknagel effect completed) gives values about 0-15 per cent.higher than the gravimetric method, while the method carried out on milk samples that have been warmed and cooled (i.e., with no Recknagel effect) gives values about 0.03 per cent. lower than the gravimetric value. The difference between the two hydrometer methods is thus about 0.2 per cent. of solids-not-fat. Slight differences can be found according to the method used for drying in the gravimetric deter- mination of total solids, but the general picture is not affected, the differences being only of the order of about 0-02 per cent.of total solids. It appears, therefore, that the ordinary Richmond method, usually carried out at about 60” F, gives results that in cold winter weather may be as much as 0-2 per cent. higher than those by the same method in warm summer weather. The Richmond formula- Total solids = - + - + 0-14508 DAVIS THE CHEMICAL COMPOSITION OF [Vol. 77 The British Standards Institution in 1938 published a method that attempted to standardise the technique in respect of the Recknagel effect and there has been much con- fusion over the two methods. Total solids = !? + 1.21F + 0.66 4 is really the same as the Richmond formula, the only adjustment made being that for the difference in the specific gravity, the B.S.I.method using density (grams per ml) at 20" C whilst the Richmond method uses specific gravity (60" F/6Oo F). Although the B.S.I. method involves warming to 40" C and cooling to 20" C in order to reverse any Recknagel effect that may have taken place, the method is not really concerned with the Recknagel effect as such. It follows, therefore, that both the B.S.I. and the Richmond methods will give virtually identical results if made on milks in which no solidification of fat has taken place, but that if the methods are used on a milk in which the Recknagel effect has been completed, Richmond's method will give a result that is about 0.2 per cent.higher than the B.S.I. method for solids-not-fat, not because of the difference in the formula for the B.S.I. method but because the warming and cooling, which is an essential part of the B.S.I. method, reverses the Recknagel effect and so lowers the specific gravity. If, therefore, a laboratory changes from the old Richmond method to the new B.S.I. method it may be expected that more accurate results will be obtained throughout the year, the general level being roughly 0.05 per cent. lower than the true gravimetric value. In cold winter weather, however, results obtained with the B.S.I. method can be as much as 0.2 per cent. lower than they would have been had the Richmond method continued in use. Hence a laboratory that uses the ordinary R5chmond method will record solids-not-fat values up to 0-2 per cent.higher in winter than in summer, and up to 0.15 per cent. higher than the true (gravimetric) result. Thus low solids-no-fat values reported by the Richmond lactometer in winter would be even lower if the -B.S.I. or gravimetric method were used. The B.S.I. formula- OTHER INVESTIGATIONS IN ENGLAND Fat and solids-not-fat figures as determined by creameries in England have been examined and reported on by Provan and Jenkins.2 Results for the period 1923 to 1946 indicate a fall of about 0.3 per cent. (8-9 to 8.6 per cent.) f o s solids-not-fat and a fall of about 0.05 per cent. (3.60 to 3.55 per cent.) in the Midlands area. Bartlett and Kay1' consider that since 1920 fat has fallen from 3.75 to 3.55 per cent.(a 6 per cent. decrease) and solids-not-fat from 8.90 to 8.65 per cent. or less (a 3 per cent. decrease). Hence, although Provan and Jenkins2 suggest a serious fall in solids-not-fat and only a slight fall in fat, the fall in fat as a percentage of the f a t itself is the more serious aspect of compositional deterioration if the figures of Bartlett and Kay are accepted. CONSIDERATION OF DATA FROM PUBLIC ANALYSTS Through the most helpful co-operation of Public Analysts, not only in England and Wales, but also in Scotland, I have been able to obtain data covering various parts of Great Britain over the period 1900 to 1950. It will, of course, be appreciated that these results are based on samples of milk taken as delivered to the public and that there may therefore be differences between these figures and those presented by workers, such as Provan2* and Provan and Jenkins,2 who have studied the results of milk samples as delivered to creameries.In requesting these data from Public Analysts, I have been most careful to lay down certain requirements, and only data from those Public Analysts who have assured me that their figures meet these requirements have been used in this paper. These requirements were laid down as follows- That the sampling has been random as far as possible. In practice, Food Inspectors usually have a system that ensures that each supplier is sampled at certain intervals, q., monthly. It may appear that this system might attach more significance to big suppliers than to small ones. T:his is probably not as serious as might be thought at first, because a big dairy company selling milk in London, for example, might sell 100 or 1000 times as much milk as a small dairyman, but whereas the small dairyman would be sampled by only one authority, the big supplier would be sampled by a number of authorities.This system of sampling, therefore, probably gives a reasonable average of the milk supply as sold to the public.Oct., 19521 MILK BETWEEN 1900 AND 1950 509 (2) That the number of follow-up and “appeal” samples form only a negligible pro- portion of the whole. Such samples naturally tend to be low and would have the effect of bringing down the average if they formed a large proportion. However, it can easily be calculated that a few per cent. of such samples will not have any significant effect on the average and, furthermore, the proportion appears to be approximately the same for all Public Analysts.(3) That the methods of analysis have been the same throughout the whole period under consideration or, if a change has been made, that the two or more methods that have been used have been shown to give virtually identical results. The effect of selecting samples in various ways is well illustrated by some data which The annual average values were were sent to me by the Public Analyst for one authority. worked out in four groups as follows- (1) All samples. (2) Excluding appeal samples. (3) Genuine milks (including those deficient in solids-not-fat but with normal Hortvet freezing-points) . (4) As (3), but with averages derived from quarterly averages in order to compensate for unevenness in sampling throughout the year.From 1931 to 1950 there was a close agreement between the four sets of average values for both fat and solids-not-fat from the point of view of the trends in quality. It was also evident that the exclusion of appeal samples made no difference to the final average value, and differences were very much smaller than the error of the methods used. The exclusion of those deficient milks that were shown to be adulterated by the freezing-point test resulted in a value that was on the average about 0.05 per cent. higher. As might be expected, this difference was less noticeable with the fat values on account of the smaller absolute amount of fat in milk. These figures show quite clearly that if there had been differences in the methods of selecting samples by the Public Analysts whose data have been presented in this paper, such differences would not affect the general results and conclusions.There are obviously certain factors that can act as sources of error in any consideration of results of this nature. For example, morning samples are almost invariably appreciably lower in fat than the evening samples, but here again it is customary for Food Inspectors to equalise the two types of milk when the milkings are known to be sold separately. Most milk is now sold by big dairy companies and this is all mixed evening and morning milk. Since, however, this milk will find its way into the sampling system whether sold separately or incorporated in the bulk, it is unlikely that this will affect the over-all average.The effect of omitting Channel Island and known adulterated samples was shown by classified data to have very little effect on absolute values and none at all on the trend of composition. TREATMENT OF DATA- Altogether analytical figures for milks sampled under the Food and Drugs Act have been received from analysts for 32 authorities, extending over the period 1900 to 1950. Figures for the beginning of the century were relatively few, so that the data can be considered more reliable for the later years. For this reason not too much reliance should be placed upon trends indicated before 1920, but from 1930 onwards the number of data and the number of different parts of the country from which the data have been derived are sufficiently great to make the results reliable.The annual numbers of results considered for England have increased from about 360 in 1900 to over 29,000 in 1950. For Scotland the annual number of 1200 in 1907 has increased to about 5000 in 1949. When annual averages were studied it was found that the yearly variations, well recognised to be due to climatic conditions, obscured any trends, and therefore averages of 5-year periods were calculated. Only clearly- defined trends are commented on, and it is suggested that a rise or fall of 0.1 per cent. or greater is significant. WEIGHTED AND UNWEIGHTED AVERAGES- problem the question is not a straightforward one. A further important factor is the proportion of Channel Island milk. These data are given in Tables VII to XIII.It is usual in summarising data such as these to use weighted averages. In the present For example, if one analyst has obtainedTABLE VII LONDON AND HOME COUNTIES Authority No. 45 46 47 49 54 Analytical methods 4, 3 later 1 4 1 Unweighted average Weighted Selection of samples (a) (4 (4 (4 (4 values average values 7- - 7- & - - 7-7 No. Fat S.N.F. No. Fat S.N.F. No. Fat S.N.F. No. Fat S.N.F. No. Fat S.N.F. Fat S.N.F. No. Fat S.N.F. 1921-25 192 6-30 2611 3.67 8.75 1931-35 189 3.74 8.76 (4) 2563 3.59 8.68 1936-40 304 3.72 8.69 1946-50 320 3.59 8.77 170 3.55 8-78 (4) 1379 3.48 8.73 (3) 7242 3.67 8.78 1941-45 363 3.73 8.80 86 3.71 8.74 (2) Fall in fat Fall in fat Fall in fat Authority No. Analytical methods 1 Selection of samples (4 10 3 (a 1 1906-10 1911-15 1916-20 1926-30 1936-40 1946-50 1921-25 1931-35 1941-45 No.Fat S.N.F. 2957 3.62 8-92 3807 3.65 8-80 3497 3.53 8.89 2757 3.61 8.82 3576 3-63 8-68 3965 3.49 8.74 Fall in fat and S.N.F. TABLE VIII NORTH-WESTERN REGION No. Fat S.N’.F. 19,530 3-66 8.89 (4) 23,208 3.62 8.78 26,923 3.55 8.78 Fall in fat and S.N.F. 25 1 and 2 (4 & No. Fat S.N.F. 675 3-67 8-81 (1) 6956 3.48 8.69 6136 3.47 8.74 1720 3-66 8.71 (2) 4409 3.67 8.68 3.67 4753 3-61 8.72 3.60 4288 3.49 8.70 3-60 3041 3.57 8.67 3.65 1870 3.52 8-70 3.59 Fall in fat 40 1 (b) No. r------h----7 Fat S.N.F. 206 3.79 8.89 929 3.82 8.85 1029 3-78 8-80 8-72 7020 3.67 8.70 8.70 7316 3.60 8-70 8.70 4592 3.51 8.70 8.74 3404 3.59 8.68 8.75 9432 3.63 8.76 43 1 (b) r No. Fat S.N.F.1) 202 3.89 9.04 1149 3-81 8.84 1533 3.79 8.75 Fall in S.N.F. 4TABLE VIII-continued 61 2 (4 7- S.N.F. No. Fat S.N.F. Authority No. Analytical methods Selection of samples 1906-10 19 11-15 1916-20 1926-30 446 1931-35 1052 1936-40 1337 1941-45 1320 1946-50 2084 7 No. 1921-25 Authority No. Analytical methods Selection of samples 1906-10 1911-15 1921-25 1926-30 1931-35 1941-45 1916-20 1936-40 1946-90 51 1 (4 .---h--7 Fat S.N.F. 3.67 8.82 (3) 3.65 8.78 3-76 8-80 3.85 8.79 3.75 8.74 7 No. 1095 1959 2043 2297 2999 6192 6642 6885 6667 59 2 (4 v No. Fat 812 3.63 978 3.69 954 3.71 742 3.70 2 1 (4 ---h---_7 Fat S.N.F. 3-58 8.83 (2) 3.60 8-80 3-66 8.79 3-60 8-82 3-62 8.78 3.62 8-83 3.61 8-79 3.66 8.78 3.65 8.79 8.81 (4) 87 3.80 8.86 (2) 8.81 8.73 8.77 67 6 later 1 (4 & 12,785 3-61 8.81 14,912 3.67 8-88 14,336 3.67 8.98 13,995 3.73 8.91 16,087 3.77 8.83 15,046 3-71 8.81 7164 3.71 8.60 22,496 3-70 8.63 Rise in fat; rise and marked fall in S.N.F.No. Fat S.N.F. 1220 3.55 8.75 (1) TABLE IX NORTH-EASTERN REGION 42 60 1 2 (4 (4 -- No. Fat S.N.F. No. Fat S.N.F. 35 3.66 8.73 176 3.81 8.79 287 3.63 8.76 4385 3.67 8.76 Unweighted average values A Fat S.N.F. 3-61 8.81 3.67 8-88 3.65 8-95 3.69 8-86 3.65 8.83 3.69 8.81 3.70 8.75 3.65 8.74 Rise and fall in S.N.F. Unweighted average values & Fat S.N.F. 3-60 8.80 3.66 8.79 3.60 8.82 3-62 8.78 3-62 8.83 3.63 8.76 3.73 8.78 3-02 8.77 co Weighted U cn t.3 average values -7 No. Fat S.N.F. 12,785 3-61 8-81 14,912 3.67 8.88 17,293 3.66 8.97 17,802 3.77 8-88 20,636 3-71 8.84 20,118 3.70 8.81 45,256 3.63 8.74 64,908 3.61 8.72 Rise and fall in S.N.F.% w M c3 M 1: 3 Y co 0 0 Weighted average values 7-7 No. Fat S.N.F. 1959 2043 2297 2999 5192 6577 7061 11,339 3.60 3.66 3.60 3-62 3-62 3.61 3.65 3-61 8-80 8.79 8.82 8.78 8.83 8.79 8-78 8-77 z *512 DAVIS : THE CREMICAL COMPOSITION OF [Vol. 77 1000 samples from a million gallons of milk and another 100 samples from 100,000 gallons of milk then a weighted average would obviously be the method to be employed. If, however, both the 1000 samples from one authority and the 100 samples from another authority are taken from one million gallons of milk, then obviously it would be more accurate to take an unweighted average. In the present investigation iboth methods have been used and weighted and unweighted averages have been calculated. The results have been grouped in regions and on a national basis.Weighted and unweighted averages give results that agree closely. Analytical methods- CODE USED I N TABLES VII TO XI11 1 = Richmond lactometer or other gravity method and Gerber. 2 = gravimetric total solids and Gerber. 3 = gravimetric total solids and Leffman-Beam. 4 = gravimetric total solids and specific gravity. 5 = gravimetric total solids and Werner - Schmidt. 6 = specific gravity and Leffman-Beam. (a) = all samples (“follow up” and “appeal” samples less than 5 per cent. of total). (b) = excluding adulterated (Hortvet) samples. (c) = excluding samples below presumptive limits. (d) = excluding “follow up” samples. No. = number of samples. Fat = percentage of fat.S.N.F. = percentage of solids-not-fat. The small figure in brackets after solids-not-fat values indicates the number of years Only data Selection of samples- Division of columns- considered when the figures for the complete 5-year period are not available. for complete 5-year periods are included for ca1cu:lation of average values. DISCUSSION OF DATA AXD CONCLUSIONS- Table V I I (London and Home Counties)-All authorities show a fall in fat. The combined averages fail to show this fall because of the inclusion of No. 49 with a high fat value for the last period. TabZe V I I I (North-Wester?z area)-Of nine authorities, two show a fall in both fat and solids-not-fat and two a fall in solids-not-fat only. Combined averages show a fall in solids- not-fat. lable I X (North-Eastem area)-The results for these three authorities are remarkably constant, apart from one high fat value for 1941 to 1945.Table X (MidZafid area)-Of six authorities, lone shows a rise and fall in fat and one a rise and fall in solids-not-fat. The averages show a slight rise and fall in solids-not-fat. Table X I (Southern area)-Both authorities show a marked rise and fall in both fat and solids-not-fat. Table X I I (Western area)-The data for this region are insufficient to allow conclusions to be drawn. Table X I 1 (Wales)-Figures for the single Welsh authority show a fall in both fat and solids-not-fat. Table X I I I (ScotZand)-Of the two authorities concerned, one shows a fall and rise in fat and the other a rise in fat. The general conclusions are summarised below- The averages reveal a significant rise in fat.Region London . . .. North-Western . . North-Eastern . . Midland . . .. Southern . . .. Western . . .. Wales . . .. Scotland . . .. Trend .. . . Slight fall in fat. .. . . Rise antl fall in S.N.F. * . . . Constant. .. . . Rise antl slight fall in S.N.F. .. .. . . Insufficient evidence. .. . . Fall in fat and (possibly) S.N.F. .. . . Marked rise in fat. . . Marked rise and fall in fat; less marked rise and fall in S.N.F.Oct., 19521 MILK BETWEEN 1900 AND 1950 513 The combined averages for all English authorities are shown in Table XIV. These unadjusted figures indicate that the solids-not-fat values have increased from 1900 to 1921-25 and since fallen. There appears to have been an increase of about 0-25 per cent., and since 1921-25 a fall of about 0-1 per cent.The fat appears to have increased slowly up to 194145 and since then fallen. It should be remembered, however, that in these figures those areas showing a fall in composition in recent years (North-Western, Midland and Southern) have contributed a greater number of results (17 out of 28 authorities) than those regions that have shown no changes in composition. Authority No. S Analytical methods 2 TABLE X MIDLAND REGION 4 3 13 4 26 2 later 1 Selection of samples (a) (4 (4 b I - 7 - No. Fat S.N.F. No. Fat S.N.F. No. Fat S.N.F. No. Fat S.N.F. 1901-05 5167 3.61 8.61 (6) 1906-10 4349 3-53 8.70 1911-15 4355 3.59 8.70 1916-20 4345 3.63 8-69 103 3.67 8.96 (1) 1921-25 4365 3.64 8.73 672 3-70 8.97 1926-30 4400 3.60 8.80 1575 8-74 8.85 (2) 751 3.61 8.88 1931-35 4404 3.60 8.81 1629 3.67 8.88 1095 3.64 8.90 1936-40 4376 3.65 8.75 3233 3.62 8.90 895 3.71 8.79 1941-45 4358 3.62 8.72 2909 3.64 8.80 837 3.59 8.84 1946-50 4358 3.59 3.72 5434 3.68 8.82 Rise and fall in S.N.F.Authority No. 41 62 Analytical methods Selection of samples 7 No. 2901-05 1906- 10 1911-15 191 6-20 1921-25 1926-30 1931-35 1936-40 50 1941-45 303 1946-50 263 1 (4 Fat S.N.F. No. 1004 3.77 8-91 (1) 4524 3-72 8-84 4240 3.66 8.83 4932 2 (b) ---h---7 Fat S.N.F. 3.63 8-80 (1) 3.68 8.80 3-72 8-72 3.59 8.73 Rise and fall in fat Kise and slight fall in S.N.F. TABLE XI SOUTHERN REGION Authority No. 68 71 Analytical methods Selection of samples 7 NO. 1921-25 540 1926-30 976 1931-35 1424 1936-40 1244 1911-45 900 1946-50 1059 Fat S.N.G.3.68 8-92 (3) 3.60 8-98 3.76 8.90 3.98 9.03 4.01 8.92 3.81 8.89 Unweighted average values - Fat S.N.F. 3.61 8.61 3-53 8-70 3.59 8-70 3.63 8-69 3-67 8.80 3.61 8.84 3.64 8-84 3.66 8.80 3.65 8.77 3.62 8-75 1 later 2 (4 r-7 No. Fat S.N.F. 2032 3-69 8-85 2422 3.82 8.91 1645 13-92 8.90(3) 1226 3.88 8-78 (3) 1605 3.74 8-72 Unweighted average values i . ' a t F . 3-65 8.92 3.79 8-91 3-98 9.03 4.01 8.92 3-78 8.81 Kise and fall in both Rise and fall in both Rise and fall fat and S.N.F. fat and S.N.F. in both fat and S.N.F. 8333 3-63 8-76 9847 3-65 8.75 7933 3-63 8-69 11,911 3-59 8.68 Weighted average values r No. Fat S.N.F. 5167 3.61 3-61 4349 3.53 8.70 4355 3.59 8.70 4345 3.63 8.69 5037 3.65 8.75 5151 3.60 8-81 15,461 3.62 8.79 22,875 3.65 8.78 20,580 3.65 8.79 26,898 3.61 8.72 Rise and slight fall in S.N.F.Weighted average values m. 3008 3.66 8.89 3846 3-80 8.91 1244 3-98 9-03 900 4.01 8-92 2664 3-78 8.78 Rise and fall in both fat and S.N.F.TABLE XI1 WESTERN REGION AND WALES Authority No. 27 Analytical methods 1 Selection of samples 1 w No. Fat 1926-30 1936-40 483 3-77 1941-45 468 3.67 1946-50 4591 3.71 1931-35 Authority No. Analytical methods Selection of samples 1901-05 1906-10 1911-15 1916-20 1921-25 1926-30 1931-35 1936-40 1941-45 1946-50 19 12-23 30 70 Unweighted Weighted 6 later 1 6 later 1 2 1 average values average values L - 7 b - 7 - b S.N.F. No. Fat S.N.F. No. Fat S.N.F. Fat S.N.F. No. Fat S.N.F. 341 3.70 8.98 (2) 8.83 (1) 3350 3.53 8-78 (3) 8.78 (1) 863 3.65 8-85 3-65 8-85 863 3.65 8.85 8.88 6254 3.73 8.80 3.72 8-84 10,846 3.72 8.83 7- No.1200 6000 6000 6000 6000 6000 6000 6000 8600 12 5 (4 -7 Fat S .N .F. 3-64 8.84 (1) 3.59 8.77 3-56 8.84 3.55 8.83 3.59 8.79 3.56 8.80 3.60 8-77 3.60 8.73 3.76 8.75 7 No. 9959 11,193 10,266 10,284 8353 12,626 31 6 Informal A- Fat S. N. F. 3.42 8.74 (4) 3.53 8.77 3.64 8.76 3-68 8.79 3.74 8-71 3-78 8-76 SCOTLAND 31 6 Statutory r--- No. Fat S.N.F. 3724 3.44 8-71 (4) 4324 3.53 8-73 4031 3.62 8.70 3922 3.64 8.77 3753 3.72 8-74 4291 3.72 8-76 57 4 Unweighted (4 average values T--A-- 7 +7 No. Fat S.N.F. Fat S.N.F. 3.59 8.77 3.56 8.84 3.55 8.83 3-59 8.79 3.54 8.77 3.62 8.74 3.64 8.76 3.74 8.73 3.75 8.76 13,838 3.65 8-79 15 (Wales) 2 1 & No. Fat S.N.F. 1006 3.70 8-89 (2) 4564 3-76 8.80 5822 3-73 8.73 4986 3.62 8.70 ' z El rn Fall in both fat and .' S.N.F.4 0 B z =; Weighted E A 0 z 'd average values 2 - 0 No. Fat S.N.F. 6000 6000 6000 6000 21,517 20,297 20,206 20,706 16,917 3.59 3.56 3-55 3.59 3-54 3.62 3.65 3.75 3.76 0 8.77 ?I 8.84 8.83 8.79 8.77 8-75 8.76 8.73 8-76 Fall and rise in fat Rise in fat Rise in fat Rise in fat Rise in fat 2 g 4 -4Oct., 19521 MILK BETWEEN 1900 AND 1950 TABLE XIV 515 ENGLAND Period 1900-05 1906-10 1911-15 1916-20 1921-25 1926-30 1931-35 193f5-40 1946-50 194145 No. of samples 3415 651 1 21,686 27,860 34,448 43,695 61,160 93,376 90,078 139,661 Unweighted average values & Fat S.N.F. 3.61 8-61 3.51 8.72 3.60 8.78 3.65 8.79 3-68 8-87 3.65 8.83 3.68 8.84 3-69 8.82 3-71 8.7 7 3.64 8.75 Weighted average values & Fat S.N.F. 3.61 8-61 3.53 8-70 3.60 8.78 3.66 8.83 3.65 8.91 3.70 8.83 3.67 8-81 3.65 8-79 3.64 8.76 3.62 8-74 Adjusted weighted averages* ,4d j ustment r-A-, factor Fat S.N.F.1-05 3-78 9-05 1.05 3-70 9.13 1-04 3.74 9.10 1-04 3.80 9.20 1-03 3.76 9.18 1.03 3.81 9-09 1-02 3.74 8.98 1.02 3-72 8.96 1.01 3.67 8.85 nil 3-62 8.74 * See p. 517 THE EFFECT OF ADULTERATION ON TRENDS IN QUALITY The question as to whether a sample of milk is adulterated or not rests on the opinion of the Public Analyst. The freezing-point test, in general use since about 1930, permits the detection of over 2.6 per cent. of added water if a depression of 0.530” C is assumed to be the limit for genuine milks. It does not detect abstraction of fat. Between 1901 and about 1930 it was customary for Public Analysts to report as adulterated any sample of milk below 3.0 per cent.of fat and 8.5 per cent. of solids-not-fat, although prosecution normally took place only when the results were 5 per cent. or more below the legal presumptive limits, e.g., 2.8 per cent. of fat or 8.0 per cent. of solids-not-fat. Before 1901 there were no legal presumptive limits and analysts had to make the best use of their analytical results and existing knowledge. Obviously until about 1930 only a proportion of adulterated milks were detected, since a milk containing 3.7 per cent. of fat could “absorb” about 19 per cent. and a milk containing 8.8 per cent. of solids-not-fat about 4 per cent. of water without adulteration being detected. Probably at least twice the numbers detected were in fact adulterated.The criteria for adulteration in these three periods (before 1901, 1901 to about 1930 and since about 1930) have thus been somewhat different, and even if data were available for the average percentage adulteration they might be misleading if used to compare the three periods. Liverseege25 discusses trends in adulteration and gives the figures recorded in Table XV. TABLE XV PERCENTAGE OF SAMPLES ADULTERATED 1873 . . . . .. .. 1879 . . .. .. .. 1889 . . .. . . .. 1899 . . .. . . .. 1909 . , .. . . .. 1919 . . .. . . .. 1929 . . .. . . .. 1930 . . .. . . .. England and Walcs 13.8 9.9 8.6 8.3 6.3 5.4 4.8 - London 15.5 13.9 10.6 8.7 4.6 3.8 2.9 - Birmingham 40.6 20.6 12.7 7.9 7.1 4.6 5.5 3.7 Before 1901 standards of 2.75 per cent. of fat and 8.5 per cent.of solids-not-fat were in use by the Inland Revenue Department, although Public Analysts preferred a fat standard of 3 per cent. (Pearmain and Moorll), and the figures shown in Table XVI were obtained for adulteration. Data for the percentages of milk samples found adulterated are given in the extracts from the Annual Reports of the Local Government Board and (later) the Annual Reports of the Ministry of Health. For historical aspects of the difficulties in the early work on the detection of adulteration and methods of analysis, see Dyer and MitchelL26516 DAVIS : THE CHEMICAL COMPOSITION OF [Vol. 77 ADJUSTMENT OF NATIONAL CURVES- We have no accurate information about the extent of adulteration of the milk of the country as a whole over the period 1900 to 1950 but we can be reasonably sure that it was appreciable in 1900 and is negligible to-day-almost certainly less than 1.0 per cent.* Although the term adulteration has always been prominent in connection with milk for the simple reason that milk is a food that is very easily adulterated by the addition of water, yet we TABLE XVI PERCENTAGE OF ADULTERATION BEFORE 1901 Total number examined 1890 .... * . .. 11,967 1891 .. .. .. .. 12,151 1892 .. .. .. .. 13,633 1893 .. .. .. .. 15,543 1894 .. .. .. .. 16,305 1895 .. .. .. .. 18,307 Adulteration, 12.8 13.4 13.3 14.9 11.5 11.1 % can say to-day with certain knowledge that milk its sold by reputable dairies is not only one of our best foods, but also one of our purest. If we had any accurate knowledge of the extent of adulteration over this 50-year period it would be possible to adjust these figures to give values that would be a more correct representation of the quality of milk as produced by the cows of the country.As such accurate information is not available, the best we can do is to make some estimate-an estimate that, although based only on general impressions, is backed by experience and will probably give a picture that is truer than the figures without any adjustment at all. For the purpose of this argument, therefore, I have assumed a 5.0 per cent. adulteration over the period 1900 to 1910, 4:*0 per cent. over the period 1910 to 1920, 3.0 per cent. over the period 1920 to 1930, 2.0 per cent. from 1930 to 1940 and 1.0 per cent. for the period 1940 to 1945. I have assumed that milk from 1945 to 1950 contains no added water.Although this is admittedly only an estimate, it is supported broadly by figures for percentage of adulterated samples that have been given by Public Analysts. These figures show quite conclusively that adulteration has progressively decreased during the century and the very small proportion of milks that are genuinely deficient (viz., non-adulterated samples below 3 per cent. in fat or 8.5 per cent. in solids-not-fat) will not affect this main conclusion. Such figures have been published by Macdonald2’ and. in Table XVII it has been worked out, from Macdonald’s figures over the period 1942 to 1947, what would be the expected amount of added water in bulk milk. These figures show a progressive decline in exactly the same way that the Public Analysts’ percentage deficitmcy figures also decrease.More precise information can be drawn from freezing-point data for bulk milk. TABLE XVII ADULTERATION OF BULK MILK Percentage of tanker bulks Year adulterated,f 1942 . . . . .. .. 32.6 1943 . . .. .. .. 8.9 1944 . . .. . I .. 2.1 1945 . . . . .. .. 2.3 1946 . . .. . . .. 6-7 1947 .. .. .. .. 3.8 Average water content of adulterated milk 1-7 1.3 1-7 1-7 1.5 1.4 Average water content of all milk 0.55 0-12 0.04 0.04 0.10 0.05 t It is assumed that a bulk milk with a freezing-point depression of less than 0.540 is adulterated; These figures do not suggest an appreciable adnlteration of milk since 1942, but the lower figures of about 0.1 per cent. were the result of “follow-up” work based on systematic testing.* Calculated as average percentage of water in all milk sold. it would contain at least 1 per cent. of added water.Oct., 19521 MILK BETWEEN 1900 AND 1950 517 It is reasonable to assume that in the absence of this checking the average adulteration of bulk raw milk in the period 1940 to 1950 was somewhat greater than 0.5 per cent. Aschaffenburg28 has also published information on the extent of adulteration of the milk supply as a whole. His figures from freezing-point data on bulk milk indicated an over-all adulteration of 0.6 per cent. for the period 1938 to 1942. It must be remembered, however, that Aschaffenburg took a depression of 0.530" C as a limit for pure milk. This limit can ignore 2.6 per cent. of added water, so that the true extent of adulteration was certainly greater than 0.6 per cent., probably about 3.2 per cent.These findings of Aschaffenburg are broadly in agreement with the adjustments made in this paper. When we come to apply these assumed percentages of adulteration to the figures, we see that the apparent rise in the quality of milk no longer exists (see Table XIV, last column). The fat values remain fairly constant throughout the whole period, but there is a marked and steady fall in solids-not-fat from 1900 to 1950. The adjusted solids-not-fat values for the beginning of the century will appear to be high, possibly excessively high, but they are at least in line with the earlier figures reported by Barham,s InglelO and Bell and Vieth. The differences may be due to the fact that Barham, Ingle, and Bell and Vieth quoted results from samples taken (at least mainly) at the farm.TRENDS IN MILK COMPOSITION IN OTHER COUNTRIES I have endeavoured to obtain the latest information about trends in milk composition in other countries and the present position is summarised in Table XVIII. It is a remarkable fact that few countries have any precise information about trends of the composition of their milk supplies. TABLE XVIII TRENDS IS MILK COMPOSITION IN OTHER COUNTRIES Country Eire . . .. . . . . New Zealand . . . . . . Sweden . . . . . . South Africa . . .. Netherlands . . * . .. Norway . . .. .. Denmark . . .. .. U.S.A, Northern Ireland, Portugal, Canada, Australia, Germany, Switzerland, Spain Changes No change observed (HarnettZ9) Fat risen from 4.3 per cent.in 1937 to 4.51 per cent. in 1950 owing to increase in proportion of Jerseys ( Riddetso) No indication of change in fat; S.N.F. not normally tested (ThomP) Fall in fat and S.N.F. in some areas, the latter due to fall in protein. Change in breed proportions alleged responsible. Fall in fat and rise in S.N.F. in other areas ( B a k a l ~ f l ~ , ~ ~ , ~ ) Fat increased by 0.54 per cent. and S.N.F. increased by 0.38 per cent. (James) Fat increased from 3.53 (1925) to 3.83 (1950) (Norw. Off. Statist.). S.N.F. rising, but declined during war ( Morkze) No data on fat and S.N.F. Lactose fallen from 4.7 to 4.5 (anhydride) in 10 years (Mogensena7) No data available METHODS FOR IMPROVING THE CHEMICAL QUALITY OF MILK The evidence presented in this paper supports that summarised by Provan and Jenkins,2 in that there has been, at least in some areas, a decline in the solids-not-fat value, and possibly also of the fat value, of milk in recent years.It is, therefore, highly desirable that some actively constructive measures should be put into practice to combat this decline. Whilst there must be the greatest support for education and advisory work amongst dairy farmers to put them in possession of the latest information about the causes of poor milk quality, those who have been actively engaged in the dairy industry will agree that there must be a definite incentive for a farmer to produce a higher-quality milk. The most effective measure is to put into operation as soon as possible a scheme of quality payment that will make it worth while for the farmer to take the necessary steps to produce milk of the highest grade, both chemically and bacteriologically.It is an interesting fact that, although we have in this country toyed with the idea of special payments for fancy grades of milk from 1923518 DAVIS : THE CHEMICAL ~COMPOSITIOK OF [Vol. 77 onwards, we are about the only advanced dairying country in the world that has not yet put into operation a simple, straightforward and 'businesslike scheme of quality payment. As far as the chemical aspect of milk is concerned, it should be noted that the farming world appears to be only interested in the fat content of milk and to consider that quality payment should be made in terms of fat, if at all. Scientists outside the dairy industry, in contrast, when discussing the subject, put forward complicated schemes based on calories or using formulae weighted to give prominence to protein, and so on.Whilst all these schemes are scientifically most interesting, and from the nutritional point of view have much to be said for them, it must be remembered that any quality payment scheme will involve more administrative machinery and inevitable cost38 and it is therefore essential that the testing side of the scheme be made as simple as possible. After consideration of this problem for many years, the author favours a scheme in which payment varies directly with total solid^,^^,^^ for the reason that it is the solids that the cow has to put into the milk and for which the farmer has to pay, whether the feeding stuffs are home-grown or imported.Although payment should be based on the total solids content (Le., a farmer would be paid per pound of milk solids and not per gallon of milk), it is most important that the farmer should be given the fat and solids-not-fat figures separately in order that he can take the necessary steps to correct a low value in either. Certain factors, such as mastitis, may affect the solids-not-fat more than the fat, while other factors, eg., the heredity of the bull, may affect the fat more. It has been argued that, as fat and solids-not-fat are correlated, it would be sufficient to pay on a fat test. While it is true that, broadly speaking, fat and solids-not-fat are well correlated, this is not so true for individual herds, and payment on fat only would not, therefore, be as just as a payment on total solids.It is important that any quality payment scheme should go out of its way to emphasise the solids-not-fat aspect for the following reasons- The solids-not-fat aspect has been more neglected in the past than the fat aspect. Genuine milks are more frequently below 8.5 per cent. of solids-not-fat than they are below 3.0 per cent. of fat. About 70 per cent. of all the solids in milk are solids-not-fat, so that it is by far the more important part of the milk without any reference to nutritional value. While one cannot compare the nutritional value of fat with that of the solids-not-fat fraction, most nutritionists would agree that, of the two fractions, the solids-not-fat is the more important, especially for children.For example, while babies and children could live quite well on skim milk (with certain necessary additions), cream would be of far less use to them. Although fat has a much higher energy or calorie value than protein and sugar, being about 2$ times more valuable, the solids-not-fat fraction of milk provides about 15 per cent. more calories than does the fat fraction. Emphasis on low solids-not-fat figures would be of considerable value in making the dairy farmer more conscious of the dangers of sub-clinical mastitis. Whilst we cannot definitely account for the greater fall in solids-not-fat, it is by no means impossible that this is due, at least in part, to the fact that in the past there have been local quality payment systems in use by buyers in which a bonus was paid only for high percentages of fat.Before the war it was a common practice for some manufacturers to pay 0.1 pence for every extra 0.1 per cent. of fat above 4-0 per cent. Thus there has been, over many years, an incentive to produce a milk high in fat, especially in the west and south, which are the main manufacturing areas. Whilst the chemical quality of milk is not a matter of great importance to the local milk distributor, it is of great importance to the manufacturer. The importance of the constituents of milk, especially fat and casein, naturally varies with the product being made.41 It is a practice in some countries to pay according to the product-yielding value of the milk. In New Zealand, payment for milk for cheesemaking is made on a fat and casein basis.Obviously for butter and cream, only the fat content is of interest. Nevertheless, from the point of view of simplicity, and in fairness to the farmer from all points of view, payment on a total solids basis is the best all-round method.Oct., 19523 MILK BETWEEN 1900 AND 1950 519 It is naturally difficult to prove that in any set of conditions a quality payment scheme would result in an improvement in quality, and the views of those who have urged the adoption of quality payment may perhaps be biased. The following extract from a letter from Dr. F. J. T. Grigg, Director of the Dominion Laboratory, Wellington, Department of Scientific and Industrial Research, New Zealand, may be of interest- “.. . In the North Island there is no evidence of deterioration in fat or solids other than fat over the last 30 years. As previously stated, this Island is the main dairying area of New Zealand.* “In the South Island, over the last 10 years or so, there has been a deterioration in fat and a very serious deterioration in solids other than fat, particularly at Christ- church. The demand for milk for human consumption has increased very considerably over that period owing to the introduction of a school milk scheme, and also the demands of military camps during the war period. There is no system of payment on quality and the tendency has been to increase volume by the extended use of lower testing cows. There is relatively little provision made by farmers in the South Island for supplementary feeding during winter months, resulting in a subnormal plane of nutrition; and in the absence of incentive payments there is no reason to expect that the position will improve.“Summarising, the decline in the South Island is a result of economic consideration, the principal cause being the lack of incentive payments for quality. . . .” Andrew42 records that for the Wellington district the fat rose steadil? from 4.05 per cent. in 1917 to 4-27 per cent. in 1926, the solids-not-fat remaining steady at about 9.0 per cent. Griggl3 reports that in the Christchurch district the fat rose from 4.08 per cent. in 1926 to 4.37 per cent. in 1936, and then fell to 4.09 per cent. in 1946. Over the same period the solids-not-fat fell from 8-97 to 8.72 per cent.For the whole of New Zealand, fat rose from 4.3 per cent. in 1933 to 1935 to 4-51 per cent. in 1950 t o 1951, owing to the increase in the proportion of Jerseys (N.Z. Dairy Board Report, 1949-50). In the Auckland district, Riddet3O reported that the fat fell from 4.6 to 4.48 per cent. and the solids-not-fat from 9.0 to 8.75 per cent. over the period 1936 to 1950. For Australia there are no precise data available, but the best-informed opinion based on much analytical work is that, whilst there is no over-all deterioration, where milk is bought on a gallonage basis (in contrast to the butterfat basis used elsewhere), the composition soon deterioratesu THE QUESTION OF STANDARDS AND THE “APPEAL TO THE cow” TEST A study of the discussions of the Wenlock Committee will show quite clearly that the Committee was concerned not only with ensuring a supply of pure milk for the public, but also with the prevention of the unjust conviction of milk producers for adulteration.With the extent of knowledge at that time and the accuracy of analytical methods then available this was a difficult problem. Few will cavil at their decision to institute legal presumptive standards together with the right of the producer to have an “appeal to the cow” test made to prove the genuineness of milk below the presumptive standards (Sale of Milk Regulations, However, the time has come when, with the more extensive knowledge and more accurate methods available to-day, we can reconsider the whole question of standards. The “appeal to the cow” test really satisfies nobody and is the cause of endless trouble and considerable expense.Basically it is unsatisfactory because the milk from any one herd can vary from day to day, especially in the fat content, and even, to some extent, in the solids-not-fat content. The Wenlock Committee was concerned primarily with establishing the genuineness or otherwise of a sample of milk, but this is not an adequate consideration. To give a simple analogy, a purchaser of fruit and vegetables would not be satisfied with what he had bought merely because he was satisfied that they were genuine articles. When one buys a pear one wants something more than a fruit that has genuinely been produced by a pear tree- one wants an attractive article of food with all the properties associated with those of a satis- factory pear.The same reasoning should be applied to milk, and it is wrong in principle to guarantee a market and exempt a producer from responsibility merely because a poor- quality milk has been shown indisputably to come from a cow. 1901). ;E Quality payment is in force [J. G. D.].520 DAVIS : THE CHEMICAL COMPOSITION OF [Vol. 77 It is suggested, therefore, that the government should give five years notice of its intention to abandon presumptive standards and institute absolute standards. This period of five years would give owners of herds producing poor-quality milk time to improve them. On the question of standards there is much to be said for retaining the 3.0 per cent. of fat and 8.5 per cent. of solids-not-fat as absolute standards, but the problem of low solids-not-fat is now much more acute than it was in 19O1.l5J6 This is well illustrated by the figures produced by Richmond at that time showing w:hat he called values for “standard” milk at different times of the year over a period of six years.These “standard” milks had a minimum fat content varying from about 2.85 per cent. to about 2.95 per cent. in May and June-the spring flush when the yield of milk rises con- siderably and the fat is at its minimum. This general picture presented by Richmond’s data is virtually identical with the picture to-day. With solids-not-fat, however, the picture given in 1894 to 1899 is very different from that to-day. Richmond’s “standard” milks invariably had their minimum values (from about 8.3 to about 8-45) in July and August, there being no other minimum in the curve.During the last ten years this seasonal variation of solids-not-fat for milks as a whole has changed considerably. The percentages of herd samples below 8.5 per cent. of solids-not-fat at the present time show two very clearly marked peaks, one in late winter and early spring and the other in late summer corresponding to the minimum in Richmond’s curves. There is little doubt, therefore, that whilst an absolute standard of 3.0 per cent. of fat would cause no undue hardship, an absolute standard of 8.5 per cent. of solids-not-fat would get many producers into trouble during March to May and July to September. To overcome this difficulty two alternatives may be considered; one would be to have a temporary absolute standard of 8.3 per cent. of solids-not-fat for a further period of five years, and the other would be to have absolute standards of 11.5 per cent.of total solids and 3 per cent. of fat. This would mean that if a farm milk fell below 8’6 per cent. of solids-not-fat, provided the fat was reasonably good, the milk would pass the test. For example, a milk of 8.2 per cent. of solids-not-fat (which is an extreme case of genuine low solids-not-fat not frequently met with) would pass an absolute standard of 11.5 per cent. of total solids provided the fat content was at least 3.3 per cent., which is not very high. It is of interest to note in respect of this latter suggestion that Lord Wenlock himself personally favoured a standard of 12-0 per cent.for total solids together with a standard of 3.0 per cent. for fat. One further suggestion is that all statutory tests should be made on mixed milks, Le., the whole of the evening milk combined with the whole of the morning milk. In practice buyers purchase both evening and morning milk together and it is unfair that any farmer should be penalised on account of the low quality of his morning milk without consideration of any compensation derived from the richer quality of his evening milk. If absolute standards were adopted there would no longer be any argument about the genuineness of milk and much unsatisfactory legislation and many time-consuming investiga- tions and prosecutions would be avoided. A satisfactory procedure in my opinion would be the imposition of a nominal fine for the sale of genuine milk below these standards and the imposition of a heavy fine when milk below the standards was shown to be adulterated, especially if the adulteration had been persistent. The present procedure has long been recognised as tending to bring the law into disrepute.The “appeal to the cow” test would, of course, cease to exist as far as the standards themselves were concerned, but an “appeal to the cow” test should be allowed for freezing- point test purposes. It is now recognised that on rare occasions a herd supply can have a freezing-point depression smaller than that hitherto a c ~ e p t e d . ~ ~ ~ ~ ~ Under this new system the only possible chance for a producer to be unfairly convicted would be if his herd happened, on one of these rare occasions, to give a genuine milk with an abnormally low freezing-point depression.In such rare cases an “appeal to the cow” test would readily settle the matter. The opinion has been expressed that, whilst such a quality payment scheme is logical and just, the complicated administrative requirements and expense make it impracticable. An alternative would be to standardise a “National Milk” of, say, 8.7 per cent. of solids- not-fat and 3.7 per cent. of fat and ask farmers to breed to conform to this standard. Any farmer whose milks failed on average throughout the year to reach these figures would be subject to a graded penalty. Due allowance for seasonal variations could be made and such a scheme would be compatible with absolute standards.Oct., 19521 MILK BETWEEN 1900 AND 1950 521 It gives me great pleasure to express my thanks to the following Public Analysts and others who have collaborated so fully with me in the preparation of this paper.Not only have they sent me all the data that were available, but they have given me all the information for which I have asked in respect of these data. Messrs. A. Alcock, D. J. T. Bagnall, H. H. Bagnall, F. C. Bullock, C. H. Chalmers, J. F. Clark, A. Dargie, A. P. Davson, S. Dixon, Dr. J. Hawthorne, Messrs. T. R. Hodgson, A. Houlbrooke, E. T. Illing, Dr. G. V. James, Messrs. A. R. Jamieson, R. Mallinder, C. W. McHugo, T. McLachlan, D. D. Moir, F. Morris, F. H. Myers, W. M. Paulley, Dr. A. Scott- Dodd, Messrs. C. J. H. Stock, R. W.Sutton, Dr. G. H. Walker, Messrs. R. W. Watridge, E. G. Whittle, A. L. Williams and A. Wilson. I am also indebted to many others who have helped me to obtain information, especially Dr. Joseph Edwards, Professor W. Riddet, Mr. W. R. Muir, Monsieur G. Mocquot, Dr. S. J. Rowland and those to whom acknowledgment is made in the references for personal communications. My thanks are specially due to Miss Catherine Gilgallon for her assistance in the treatment of the analytical figures. 1. 2. 3. 4. 5 . 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. REFERENCES Edwards, J., J . Fmrs.’ Cl., 1950, No. 6, 80. Provan, A. L., and Jenkins, D. I., J . Soc. Dairy Tech., 1949, 2, 88.Kay, H. D., J . Roy. San. Inst., 1947, 6, 515. Davis, J. G., McClemont J., and Rogers, H. J., J . Dairy Bes., 1939, 10, 59. Stableforth, A. W., J . Minist. Agric., 1935, 61, 945. Foot, A. S., and Shattock, P. M. F., J . Dairy Res., 1938, 9, 166. Rowland, S. J., and Zein-el-Dine, M., Ibid., 1938, 9, 182. Barham, G., J . Brit. Dairy Fmrs.’ Ass., 1885, 2, 48. Richmond, H. D., “Dairy Chemistry,” Charles Griffin & Co., Ltd., London, First to Fifth Editions, Ingle, H., “Report on the Solids in Cows’ Milk,” Yorkshire College, Leeds, 1900. Pearmain, T. H., and Moor, C. G., “The Analysis of Foods and Drugs,” Baillibre, Tindall and COX, Cranfield, H. T., J . Agric. Sci., 1927, 17, 62. Burr, H., M.Sc. Thesis, Reading University, 1939. Rowland, S. J., J .Dairy Res., 1944, 13, 261. -, Dairy Ind., 1946, 11, 656. -, Chew. Ind., 1948, 106. Bartlett, S., and Kay, H. D., J . Roy. Agric. SOC., 1950, 111, 87. Featherstone, J., Rickaby, C. D., and Cavell, A. J., J . Dairy Res., 1951, 18, 155. Board of Agriculture, Rep. Dept. Ctte. Regulations for Milk and Cream (Cd. 491) H.M. Stationery Tocher, J. F., “Variations in the composition of Milk,” H.M. Stationery Office, London, 1925. Hanley, J., “Progress in Milk Technology,” J. Bibby & Sons, Ltd., Liverpool, 1936. Boden, S. M., and Campbell, C. H., J . Dairy Res., 1942, 13, 45. Rowland, S. J., Personal communication, 1951. Provan, A. L., Int. Dairy Congr., 1949, 2, 217. Liverseege, J. F., “Adulteration and Analysis of Food and Drugs,” J. & A. Churchill, Ltd., Dyer, B., and Mitchell, C.A., “Fifty Years of the Society of Public Analysts,” W. Heffer & Sons, Macdonald, F. J., Analyst, 1948, 73, 423. Aschaffenburg, R. A., J . Dairy Res., 1944, 13, 281. Harnett, P., Personal communication, 1951. Riddet, W., Personal communication, 1951. Thom6, K. E., Personal communication, 1951. Bakalor, S., Department of Agriculture, S. Africa, Bull. No. 287, 1947. -, Department of Agriculture, S. Africa, Bulls. Nos. 285 and 301, 1948. -, Personal communication, 1951. Janse, L. C., Neth. Milk Dairy J., 1950, 4, 1. Mork, R., Personal communication, 1951. Mogensen, M. S. , Personal communication, 1951. Davis, J. G., J . SOC. Dairy Tech., 1948, 1, 81. -, Dairy Ind., 1945, 10, 27. -, Dairy Farmer, 1951, August, p. 19 and September, p. 15. - “Dictionary of Dairying,” Leonard Hill, Ltd., London, 1950.Andrew, R. L., N.Z. J . Sci. Tech., 1928, 10, 229. Grigg, F. J. T., N.Z. J . Sci. Tech., 1947, 29, 101. Loftus-Hills, G., Personal communication, 1951. Hillman, H. C.. Provan, A. L., and Steane, E., Chem. & Ind., 1950, 333. Sykes, D. A., Macdonald, F. J., and Davis, J. G., Unpublished observations, 1946. 1899, 1914, 1920, 1942, 1952. London, 1897. Office, London, 1901. London, 1932. Ltd., Cambridge, 1932. EXPRESS DAIRY COMPANY LIMITED LONDON, W.C.l TAVISTOCK PLACE,,522 DAVIS : THE CHEMICAL COMPOSITION OF [Vol. 77 DISCUSSION THE PRESIDENT said that this was a most important subject, and not many people were in a position to deal with it. The Society was, therefore, fortunate in having Dr. Davis present this paper for discussion.There were many factors that might bring about a change in quality of milk, such as the considerable increase in the numbers of Friesian cows. Farmers had always been concerned with fat content, and an increase in fat content was often accompanied by a slight increase in solids-not-fat. An important difference between present-day practice and that of the earlier quarter of this century was that almost all milk was now artificially cooled; the fact that coolers had a tendency to leak slightly on occasion might lead to a small reduction in the percentage of solids-not-fat. DR. N. J. BERRIDGE pointed ont the possibility of condensation of atmospheric moisture on excessively cooled milk. DR. DAVIS said that the extent of adulteration from cortdensation of atmospheric moisture on excessively cooled milk was probably of the same order as that due to residual rinsing water in utensils, and would not be a highly significant factor.DR. E. C. WOOD said that i t must surely make a difference whether the Public Analyst received all the samples of milk taken by the sampling officers, or whether they “sorted” the samples and sent only the suspicious ones to the analyst. He added that, in his experience, there were certainly more samples deficient in non-fatty solids than in fat, if all the samples were considered; but, if one removed the samples :shown by the Hortvet test to be watered, there remained more fat deficiencies than non-fatty solids deficiexces. Finally, he asked if Dr. Davis had noticed an apparent negative correlation in some of his curves, between fat content and non-fatty solids content over a period of years.DR. DAVIS said that care had been taken to make sure that only a very small proportion of “follow up” samples had been considered in computing the average values. This probably was always less than 5 per cent., and would not have any significant effect, and further, it appeared to be approximately the same for all authorities. Dr. Wood’s point about differences in the nature of deficiencies in genuine and watered samples was reasonable, if one remembered that watering would reduce fat and solids-not-fat, whereas in genuine samples a common source of error was failure to keep fat properly distributed. Other things being equal there would be more chance of a milk being deficient in solids-not-fat than in fat because of the relation between average values and the legal presumptive limits.There is always some correlation between fat and solids-not-fat values because the greatest change in the monthly average values takes place when the cows go out to grass. There is always a marked increase in yield accompanied by a rise in solids-not-fat and a fall in fat. MR. C. W. MCHUGO said that there was one point with which Dr. Davis had not dealt in his very comprehensive review. This was the well-known fact that morning milkings, especially a t certain periods of the year, were liable to be low in fat, and it was by no means unusual to find that the fat content of such milks was appreciably below the Statutory minimum of 3.0 per cent.He was aware that the large dairy companies overcame this difficulty by bulking together the morning and evening milkings, but in certain areas where producers had not the necessary facilities for doing this, morning milkings were being sold direct to the public. He felt, therefore, that the authorities responsible for the distkibution of milk should take what steps might be necessary to see that morning and evening milkings were properly blended, and so ensure that the public received milk of at least average good quality. The Government had recently set up a Working Party on Quality Milk Production, and he trusted that this important aspect of the subject would receive their attention. DR. DAVIS agreed with the importance of this point but in his opinion a milk producer should be judged on the mixed meals of the morning and evening milking, and not on the morning milk alone.Most milk was now bulked and the public nearly always got the mixed morning and evening milk. MR. N. HERON said that he had experienced in the north of England a strong suggestion of widespread adulteration of a minor nature. Large batches of milk were encountered with a remarkably constant set of figures, viz., 3.3 & 0.2 per cent. of fat, 8.6 f 0-05 per cent. of solids-not-fat and a freezing-point range of 0.528 to 0.535, which indicated about 2 per cent. of added water. DR. DAVIS thought that the consistency of the analytical figures might be due ‘to controlled adultera- tion, but it must be borne in mind that when milk is bulked the composition tends to be very consistent for any one area.Her experience was that bottled milk had shown a steady, although slight, decline since 1937 in the solids-not-fat content, rather than in fat content. He wondered if Dr. Davis had taken this into account. MISS B. M. FISCHER supported the author’s contention that the quality of milk was declining. The averages for bottled milk were as follows- Fat, yo Solids-not-fat, yo November, 1937 . . .. . . 3.86 8-72 November, 1960 . . .. . . 3-9 8-69 April, 1938 . . . . . . . . 3.46 8.70 April, 1951 . . .. .. . . 3.62 8.48Oct., 19521 MILK BETWEEN 1900 AND 1950 523 When individual producers were considered, the fall was even greater, as shown by the following figures- 1944-1945 .. .. . . . . 3.6 8.38 1950-1951 .... .. . . 3.35 8.16 with fourteen out of fifteen samples below 8.5 per cent. of solids-not-fat and one below 3 per cent. of fat in 1950-1951. That the fall in quality of bulk milk was not more serious was due to the fact that while many producers’ supplies had deteriorated or remained consistently poor, others had maintained or even improved the composition of their milks. The proposal to pay for milk on quality was supported as the only effective measure to improve the situation as a whole. Some farmers were already taking steps to improve the quality of their milk in anticipation of quality payment. DR. J. H. HAMENCE said that he was delighted to find, a t last, a man after his own heart who supported his suggestion of an absolute standard for milk.He believed that this was the only method by which the quality of milk would ever be maintained and that, unless some rigid deterrent was provided, the standard of milk would gradually go down and down. He was disappointed that Dr. Davis had not given more information as to the proportion of abnormal milks, that is, the proportion of milks falling below the standards laid down in the Sale of Milk regulations. As a Public Analyst, Dr. Hamence was seriously concerned a t the big increase in abnormal milks over the last 20 years, and this applied to all the different parts of England for which he acted. Whereas in 1930 it was rare to find appeal-to-cow samples that contained less than 8.5 per cent. of non-fatty solids and less than 3 per cent. of fat, a substantial proportion of appeal-to-cow samples examined to-day were below the standard, and some of them had dropped to very low levels indeed.He felt that, if milk were sold to the consumer at a price based on quality, there would be a general tendency to buy the lower-priced article, and this would, in turn, tend to encourage the production of poorer milk. Dr. Hamence realised that there would be many difficulties in the way of establishing an absolute standard, but he was of the opinion that it was the only way to stop the fall in quality. DR. DAVIS pointed out that the question of the proportion of abnormal milks containing less than 3 per cent. of fat or 8-5 per cent. of solids-not-fat had been dealt with by Dr. S . J. Rowland and other workers in this field.It was true that a t certain times of the year, especially in the late winter and early spring, the solids-not-fat content of a large proportion of farmers’ milks could fall below 8-5 per cent. He had known of some creameries where this proportion was, on occasion, as high as 50 per cent. It was generally agreed by those who had studied the subject, that the most important factor was the difficulty of feeding-stuff s, although the increase in winter production, the increase in the proportion of Friesian cattle and the general breeding policy were also significant factors. It was comparatively rare for farmers’ milks to have a fat content of less than 3 per cent. MR. F. N. GINGELL agreed with the author’s findings of a downward trend in quality, but thought that this was due to differences in breed.At one time a Friesian herd was unusual: now there were more Friesians than Ayrshires. The English breeds of big-framed cows gave milk with the higher solids-not-fat contents. He suggested that the fall in quality was not due to any emphasis on quantity; the increase in the quantity of milk produced had come from an increase in the number of dairy herds. DR. G. H. WALKER asked why, if the decrease in solids-not-fat was caused by an increase in the number of Friesian herds, there was not a similar decrease of fat content. In Lancashire a marked drop in solids- not-fat had occurred since 1941 rather than a slow drop over the last 20 to 30 years, and he asked whether any other districts had shown a similar trend. DR. DAVIS said that the fall in solids-not-fat being more noticeable than the fall in fat was in agreement with the theory that the major factor responsible for the decline in milk quality was feeding difficulties.MR. J. R. ROWLING agreed with Dr. Davis about the downward trend in quality, but asked whether, in considering average figures taken from Public Analysts, allowance had been made for the segregation of Channel Island milks into the premium market. This could lead to the Public Analyst being called upon to analyse a smaller proportion of Jersey milk samples in comparison with samples of ordinary milks, and this could colour the figures quoted by Dr. Davis. DR. DAVIS said that he had not overlooked this point; however, it could readily be calculated that irrespective of whether Channel Island milk was incorporated in the bulk or sold as such, the effect on the average of a large number of Public Analysts’ samples would be the same.Nevertheless, the proportion of Channel Island milk was very small and would be unlikely to affect the over-all average. MR. R. W. SUTTON did not agree with Dr. Hamence that unless something was done, quality would go down and down. Clearly this could only be attained either by an increased cow population or by an increased yield per cow, and in this country, with its diminishing area of agricultural land, the second factor had necessarily received greater attention. There had been a swing over to the breeds with higher yields and, in his opinion, this accounted for the changes in average composition that had been recorded.In all the breeding for production, there did not appear to be much evidence of any decline in the average figures for fat and non-fatty-solids within any particular breed. Fat, % Solids-not-fat, ”/o A few years ago there was a drive towards increased milk production.DAVIS [Vol. 77 The changes that had been recorded appeared to be variable. In the Derbyshire area they were not marked. The average fat content had fallen from 3-63 per cent. in 1935 to 3-59 per cent. in 1949, but during this interval had been as high as 3.77 per cent. and as low ils 3.66 per cent. The average figure for non-fatty- solids had fallen from 8.80 in 1936 to 8.72 in 1949, and had ranged from 8-83 to 8.67 in the intervening years. Payment on a quality basis was now advocated, but this raised the very great question as to how quality should be assessed. Determinations of percentages of fat and milk solids would naturally give a measure of manufacturing quality, but as a large proportion of the milk produced in this country was sold for liquid consumption, nutritional quality and hygienic quality ought surely to receive consideration. Dr. Davis had shown a table to indicate that milk could be a most important food in supplying our daily requirements of calcium, vitamin A and riboflavin and other members of the vitamin-B complex. Unless it could be shown that there was a good positive correlation between these factors on the one hand, and the percentages of fat or milk solids on the other, it was clear that no simple tests were available for the reliable assessment of nutritional quality. The assessment of hygienic quality required separate tests, and might be considered as outside the scope of the present paper; but i t was reasonable to suppose that the average consumer of liquid milk would expect some at tention to be paid to this factor in any attempt to measure quality. Since there were daily variations in the quality of milk from any herd andl since there were variations in the composition of the herds throughout the year, many samples would need to be examined. Even if the examination were limited to the simple determinations of fat or milk solids, it was clear that the cost of the work would be great, and if to this is added the recognition that the tests might still not provide a true measure of quality, it seemed to him that the cost would be out of all proportion to the value of the work. Finally, it seemed to him that a limited application of these proposals was available to the Dairy Companies a t present. In a report on “The Cost of Milk Distribution” that had been published in 1940, there was evidence of very considerable variation, the co:jt of distributing a gallon of milk ranging from sixpence or less, a t one extreme, to a shilling or more at the other. Even if this gap had been closed to some extent, it was likely that differences would still be .found, and since the distributors’ margin was a level one and since the larger Dairy Companies would be counted as the most efficient, it seemed to him that there might be ample margin a t present for the large companies to pay a bonus for quality, and this could quite properly be based entirely on their own assessment of quality. He had given considerable thought to the question of criteria for quality payment, and he was strongly of the opinion that the criterion must be a very simple one and the same for all producers. From the chemical and nutritional aspects he considered that the total solids was the best. A dye reduction test would be adequate for bacteriological quality, bearing in mind that only the average of a number of results would be used as a basis. He did not agree that the cost of testing for a quality payment scheme would be out of all proportion to the value of the work. Buyers were already testing farmers’ milk for chemical and bacteriological quality a t least once a month. The present system of testing, w!hich was practically the same throughout the country, could, with suitable modifications, become the basis of a quality payment scheme. He did not feel that he could deal with Mr. Sutton’s last point since the question of distributors’ margins and increases in costs of handling milk was a very controversial one. He would merely say that before the War manufacturing firms frequently paid an extra premium for good chemical quality, and that some buyers of milk that was used for liquid consumption paid a premium for good bacteriological quality. The number that paid premiums in present circumstances was extremely small. There were, he understood, about 160,000 registered milk producers in the country. DR. DAVIS thanked Mr. Sutton for raising some very important questions.

ISSN:0003-2654    

DOI:10.1039/AN9527700499

出版商:RSC    

年代:1952

数据来源: RSC

摘要:

524 DAVIS [Vol. 77 ERRATUM : August (1952) issue, p. 419. On the 16th line from the bottom, for “ ferrous ” read ‘ I cuprous.”

ISSN:0003-2654    

DOI:10.1039/AN9527700524

出版商:RSC    

年代:1952

数据来源: RSC

摘要:

Oct., 19521 COOK AND ROOKS 526 A Nomogram for the Calculation of Urinary 17-Ke t 0s ter oids BY E. R. COOK AND MARGARET E. ROOKS If 24-hour urine specimens are diluted to multiples of a specified volume, the total urinary 17-ketosteroids can be calculated without loss of accuracy by the use of a nomogram, the construction of which is described. The graph can be used for several different methods by taking suitable volumes of urine and solvent. The accuracy of the correction by equation and by nomogram has been tested, and the close agreement of corrected values obtained by the procedures of Callow, Callow and Emmens (1938), Hamburger (1948) and Cook (1952) is shown. URINARY 17-ketosteroids are determined by hydrolysis of the specimen with hydrochloric or sulphuric acid and extraction with an organic solvent, followed by colorimetric or polaro- graphic estimation of the washed and dried extract.The crude extracts generally contain non-specific chromogenic material that does not interfere with the polarographic procedure of Barnett et aZ.,lJ but which intensifies the colour produced in the widely-used Zimmerman determination, sometimes increasing the results to twice their true value. This interference can be eliminated by the Girard-T chemical separation, as described by Girard and Sandule~co,~ Talbot, Butler and MacLachlan,4 Pincus and Pearlman5 and Cook,G or by use of a correction factor. This paper describes such a factor and its application in a graphical method for the calculation both of crude total 17-ketosteroids and of the corrected values in mg per 24 hours.The nomogram has been applied to three different methods for the estimation of urinary 17-ketosteroids, and the results compared. CORRECTION FACTOR The equation depends on the difference in the absorption maxima shown by the interfering chromogens and 17-ketosteroids in the anhydrous Zimmerman colour reaction, which makes use of alcoholic potassium hydroxide and nz-dinitrobenzene, as described by Zimmerman? and Callow, Callow and Emmens.* The chromogens show a broad absorption maximum in the violet, and the 17-ketosteroids have a sharp maximum in the green. Detailed accounts have been published by Gibson and E ~ e l y n , ~ Fraser, Forbes, Albright, Sulkowitch and Reifenstein,lo Werner,ll Talbot, Berman and MacLachlan,12 Engstrom and Mason,13 Hamburger14 and Allen.15 The equation used in this procedure is- (Kc x Eg) - E v Corrected Eg = Kc - Kp ' where E, = extinction with Ilford No.604 Spectrum green filters of the sample under test, E, = extinction with Ilford No. 601 Spectrum violet filter of the sample under test, K, = E,/Eg of the non-specific chromogenic material, and K, = E,/Eg of pure 17-ketosteroids. In order to determine Kc, a large-scale Girard-T separation was carried out on 600 mg of crude 17-ketosteroids, which were obtained by combining the benzene extracts from more than 300 24-hour specimens. The E,/E, of the non-ketonic fraction so prepared was 1.20, while the E,/Eg of pure androsterone was 0-36. These values must be established by each individual laboratory. The correction factor then reduces to- 1.20 Eg - E, 0.84 ' Corrected Eg = The use of this equation assumes that the value of E,/Eg for the non-specific chromogenic material remains constant for all specimens received.This is not correct, as there is some526 COOK AND IiOOKS: A NOMOGRAM FOR THE CALCULATION slight variation in this value between individuals, and even from day to day in the same individual, but for clinical purposes the effects 011 the calculated result is negligible. [Vol. 77 CONSTRUCTION AND us13 OF NOMOGRAM By use of the correction factor described above, a nomogram, Fig. 1, was constructed, based on that of Hamburger,lG by which the corrected E, (A) can be found for any sample for which values of E, (B) and E, (C) have been determined. By superimposing a standard (C). Observed E, Fig.1 A, B, C: nomogram relating observed :E, and E, to the corrected E,. A, D, E : standard graph of androsterone or dehydroisoandrosterone. Scales F: volume scales relating the corrected weight (in mg) of 17-ketosteroid to the 24-hour value curve (A, D, E), drawn for androsterone or dehydroisoandrosterone as reference substance, it is possible to read directly the weight in milligrams of 17-ketosteroid equivalent to the corrected Eg, and to calculate the result in mg per 24 hours by the scales (F). The principle of these scales is simple and of general application, and by appropriate choice of volumes of urine and solvent, calculations for several methods can be made on the same set of scales. In the present study three different procedures have been compared, and in each the 24-hour urine specimen was diluted to the nearest multiple of 400ml with distilled water.Method I (CaZZow, Callow and Emmens8)-A solution consisting of 240ml of diluted urine and 37 ml of concentrated hydrochloric acid was heated under reflux for 1 hour with 45 ml of carbon tetrachloride; this procedure was repeated twice with two further 45-ml portions of carbon tetrachloride. The extract was washed with 2 N sodium hydroxide and then with water, and was diluted to 150 ml with carbon tetrachloride; 5-ml aliquots of this solution were evaporated to dryness for the colorimetric determination. Method I1 (Hamburger16)-A solution consisting of 40 ml of diluted urine, 4 ml of 40 per cent. v/v sulphuric acid and 40 ml of benzene wat'j heated under reflux for 35 minutes.The washed extract was diluted to 50 ml with benzene, and 10-ml aliquots were taken for the estimation. Method 111 (Hambwurger,lG modijed by CookG)--A solution consisting of 240 ml of diluted urine and 24 ml of 40 per cent. v/v sulphuric acid was heated under reflux for 35 minutes with 45 ml of benzene; this procedure was repeated twice with two further 45-ml portions ofOct., 19521 OF URIXARY 17-KETOSTEROIDS 527 benzene. The washed extract was diluted to 150ml with benzene and 5-ml aliquots were taken for the estimation. The calculation for Methods I and I11 then becomes- Ketosteroid, mg per 24 hours = corrected weight of ketosteroid x - 150 400 x M 240 5 = corrected weight of ketosteroid x 50M, and for Method I1 it becomes-- 50 400 x M Ketosteroid, mg per 24 hours = corrected weight of ketosteroid x - x - 10 40 = corrected weight of ketosteroid x 50M, where M = 1, 2, 3, and so on, and all weights are in milligrams.The scales (F) are calculated in milligrams of androsterone x 50 M, and give the corrected 17-ketosteroids in mg per 24 hours corresponding to the total diluted volume of 24-hour urine and the determined E, and E,. A 24-hour urine specimen was diluted to 1600 ml and the determination carried out. The extinction readings were E, = 0.473 and E, = 0.325. The crude 17-ketosteroid present is found by drawing a vertical line from 0-473 on (A) to line (D) and then drawing a horizontal line from this point of intersection to the 1600-ml scale (F). The corrected value is found by drawing a line from 0.473 on scale (C) through 0.325 on (B) to the point of intersection on (A), which gives the corrected E,.A vertical line drawn from this point to the curve (D), then a horizontal line drawn from (D) to the 1600-ml scale gives 13.5 rng per 24 hours, which is the “true” value. The standard androsterone curve should be checked frequently, but since all other scales remain fixed, it is a simple matter to erase curve (D) and substitute the new standard. An example will make this clear. The result is 22.6 mg per 24 hours. RESULTS To check the accuracy of the correction factor, 240-ml aliquots from the urines of 50 patients were hydrolysed and extracted by the modified Hamburger procedure (Method 111). The Zimmerman reaction was carried out on the extract, and the extinctions with Ilford Spectrum green (No.604) and Spectrum violet (No. 601) filters were used to calculate the corrected 17-ketosteroid values. The ketonic material in the crude extract was obtained by a micro-scale Girard-T separation.6 Table I shows the close agreement between results corrected by chemical separation and those corrected with the formula. These results TABLE I k C U R A C Y OF CORRECTION FORMULA 17-Ketosteroids, mg per 24 hours Uncorrected chemical reparation formula A 7 - 7 Corrected by Corrected by 41.0 36.5 30.5 22.2 16.8 16-1 2 1 4 17.4 16-4 20.6 13.1 12.5 18-4 14.6 14.7 16.8 11.7 11.4 15.8 12.9 12-3 14.6 8.3 7-8 13.6 10.0 9.5 10.3 8.6 7.5 are comparable with those of Morris17 and Butt, Morris, Robinson and Warren,lS who showed that results by polarographic methods, in which the non-specific chromogens are not estimated, are in good agreement with those by colour correction methods.The 17-ketosteroid values found by chemical separation are generally slightly higher than those obtained by calculation, probably because of the small amount of residual non-specific chromogenic material that the Girard-T separation fails to remove. Table I1 was constructed with hypothetical values of E, and Ev, and shows the close agreement between values found graphically and by calculation.528 COOK AND ROOKS: A NOMOGRAM FOR THE CALCULATION TABLE I1 COMPARISON OF CORRECTIONS BY FORMULA AND BY GRAPH Extinction values A I -l =, E v 0.10 0.04 0-06 0.08 0.12 0.16 0.20 ,Om08 0.40 0.60 0.80 0.16 0.24 0.32 0.24 0.36 0.48 0.32 0.48 0.64 Corrected E, r A \ By formula BY graph 0.095 0.094 0.07 1 0.068 0.047 0.043 0.190 0.192 0.143 0.142 0.095 0.093 0.382 0.286 0.191 0.571 0.429 0.286 0.763 0.572 0-381 0.386 0.286 0.186 0.576 0.430 0.281 0.765 0.570 0.375 [Vol.77 Table I11 shows the similarity between the results of the three extraction procedures described above. The temperature of the urine during the simultaneous hydrolysis and extraction with benzene is approximately 91" C, while with carbon tetrachloride it is only 68°C. This difference may account for the smaller amount of non-specific chromogenic material found in carbon tetrachloride extracts, which results in lower uncorrected values than those found in the benzene extraction.TABLE 111 Sampie 1 ? - > 4 5 6 7 8 9 10 11 12 13 14 15 COMPARISON OF EXTRACTION METHODS 17-Ketosteroitds, mg per 24 hours Method I Mlethod 11 Method 111 7 A -l 7 bncorrected Corrected Uncorrected Corrected Uncorrected Corrected 5 1.4 45.0 50-0 45.0 - - 32.7 29.0 35-0 28.2 43-0 36.1 28.1 24.5 33.8 25.1 33.0 28.3 25.5 21.2 29.0 22.9 29-5 24.0 17.0 13.0 19-7 13.2 2 1.4 14.3 13.4 10.7 15.5 10.7 18.4 13-0 11.1 9.5 15-1 10-7 __ - 11.5 8.6 14.8 8.8 14.7 8.7 10.3 7.5 13.1 7.2 11.8 7.9 11.0 8.4 13.4 9.8 9.2 7.9 10.2 7.7 10.7 9.2 6.0 4.8 8.5 5.7 8.5 5.7 6.9 3.7 5.7 1.0 8.0 3.7 4.1 2.3 5-4 2.5 5.8 2.7 3.6 2.3 3-5 1-8 - - - - The authors gratefully acknowledge the valuable technical assistance given by Miss Jean Rex and Mr. A. E. Hall. 1 . 3. 4. 5. 6. 7. 8. > Y. REFERENCES Barnett, J., Henley, A.A., and Morris, C. J. 0. R., Biochem. J . , 1946, 40, 445. Barnett, J., Henley, A. A., Morris, C. J. 0. R., and Warren, F. L., Ibid., 1946, 40, 778. Girard, A., and Sandulesco, G., Helv. China. Actn, 1936, 19, 1095. Talbot, N. B., Butler, A. M., and MacLachlan, E. A., J . Bid. Chem., 1940, 132, 595. Pincus, G., and Pearlman, W. H., Endocrinology, 1941, 29, 413. Cook, E. R., Analyst, 1952, 77, 34. Zimmerman, W., Hoppe-Seyl. Z., 1936, 245, 47 Callow, N. H., Callow, R. K., and Emmens, C. VC'., Biochenz. J., 1938, 32, 1312.Oct., 19521 OF URINARY 17-KETOSTEROIDS 529 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. Gibson, J. G., and Evelyn, K. A., J . Clin. Invest., 1938, 17, 153. Fraser, K. W., Forbes, A. P., Albright, F., Sulkowitch, H., and Reifenstein, E. C., J . Clin. Endo- Werner, S. C., Ibid., 1941, 1, 951. Talbot, N. B., Berman, R. A., and MacLachlan, E. A., J . Biol. Chew., 1942, 143, 211. Engstrom, W. W., and Mason, H. L., Endocvin.ology, 1943, 33, 229. Hamburger, C., Acta Endocrinol., 1948, 1, If). Allen, W. M., J . Clin. Endocvinol., 1950, 10, 71. Hamburger, C., Acta Endocriizol., 1948, 1, 375. Morris, C. J. 0. R., J . Endocvinol., 1948, 5, lxxi. Butt, W. R., Morris, C. J. 0. .R., Robinson, A. M., and Warren, F. L., Ibid., 1951, 7, xii. crinol., 1941, 1, 234. BIOCHEMICAL AND ENDOCRINOLOGICAL RESEARCH DEPARTMENT BRISTOL MENTAL HOSPITALS BARROW GURNEY, BRISTOL A p d , 195%

ISSN:0003-2654    

DOI:10.1039/AN9527700525

出版商:RSC    

年代:1952

数据来源: RSC

摘要:

Oct., 19521 OF URINARY 17-KETOSTEROIDS 529 The Determination Of Small Amounts of Calcium in Plant Material BY A. C. MASON Small quantities of calcium (10 to 500 p g ) can be estimated quickly and accurately by titration with an aqueous solution of sodium ethylenediamine tetra-acetatey with Murexide as indicator. Phosphates, which interfere, are removed by ion exchange on a resin column. The design of such a column and its reliability in effecting the complete and rapid removal of phosphate is discussed. The method, which is free from interference by magnesium, iron or manganese, in amounts usually encountered, gives results that are in good agreement with the oxalate method. The standard error for 16 deter- minations on a plant ash solution is 2-84 pg in 212 pg, which corresponds to a coefficient of variation of 1.35 per cent.AN accurate method of estimating small quantities of calcium in biological material has long been required. The standard method, which depends on the precipitation of calcium oxalate, besides being somewhat laborious, ceases to be satisfactory for quantities much less than 1 mg of calcium, on account of solubility difficulties. Sodium ethylenediamine tetra-acetate (subsequently referred to as EDTA) has been used by Willsonl for the determination of calcium and magnesium in leaf tissue. He claimed satisfactory results for 4 to 8-mg quantities of calcium, but took no steps to avoid interference from phosphate ions, although he was aware of this source of error; consequently, his method is of limited application. Independently, Mason2 briefly outlined in a preliminary note a method for estimating smaller quantities of calcium (100 to 200 pg) in plant material.Early results indicated that phosphate interference could be avoided, without a separation, by adding an excess of EDTA before making the solution alkaline (so preventing formation of calcium phosphate) and subsequently titrating back with standard calcium solution. How- ever, extended experience showed the method to be unreliable for routine work and it became obvious that it is imperative to remove interfering phosphate ions. Essentially the method to be described consists in percolating the sample, after suitable dilution, through a small resin column, whereby the calcium and other cations are held in the column while the phosphate ions pass through.The calcium is then eluted with hydrochloric acid and titrated with EDTA. PRELIMINARY EXPERIMENTAL WORK DESIGN OF THE RESIN COLUMN- In order to achieve speed in operation, experiments were made to find the smallest dimensions of the resin bed consistent with complete retention of calcium. The particle size of the resin was standardised by sieving the ground resin between Nos. 60 and 120 B.S. sieves. This fairlv fine fraction was chosen to give an adequate surface area of resin without530 MASON : THE DETERMINATION OF SMALL AMOUNTS OF unduly impeding the percolation. The results of many trials indicated that retention of calcium was uncertain if the bed depth was much less than 25 mm, but the diameter of the bed could be varied considerably without effect.In all tests 500 pg of calcium, contained in 5 ml of water, were run through the freshly regenerated and washed resin column, and the effluent was titrated with EDTA. From tests carried out on columns prepared as described under Method on p. 531, it was found that there is a risk of calcium ions leaking through the column if the percolation rate exceeds 0.6 ml per minute. The resin selected for preliminary work was Zeo-Karb 215, a cation-exchange resin that contains sulphonic acid and hydroxyl group:;. The high-capacity sulphonic-acid resin Zeo-Karb 225 was also tested to see whether the percolation rate could be increased or the bed depth decreased, but the results were unsuccessful. [Vol. 77 ELUTIOK OF CALCIUM- The reaction between resin and cations in solutioii is usually represented as- RH, + Ca" + CaR.H,-2 + 2H'.When the concentration of hydrogen ions is sufficiently low the equilibrium is displaced to the right, and calcium ions are removed from solution and held by the resin (the effect of acid concentration in preventing ion exchange is shown in Table I). As the concentration of hydrogen ions is increased the reaction is displaced to the left, and calcium ions are liberated; TABLE I[ EFFECT OF ACID CONCENTRATION ON IOK-EXCHANGE EFFICIENCY 505 pg of calcium were taken in each example and mixed with hydrochloric acid, and the calcium held by the column was estimated Concentration of hydrochloric acid Calcium recovered, PI-: 1.0 N 110, 107, 90, 95 0.1 n; 502, 504, ,503, 504 0.01 N 502, 507, 607, 505 the latter process, referred to as elution, is efficient in the presence of 5 N hydrochloric acid.Concentrated hydrochloric acid attacks the resin and interferes with the final titration. Experiments were carried out to determine the minimum volume of acid for complete elution ; for 1000 pg of calcium this was established as 6 ml. There is a tendency for the last few micrograms of calcium to be held rather more tenaciously by the resin, and this is illustrated by the figures given below, in which 4 or 6 ml of 5 N hydrochloric acid were used to elute 207 pg of calcium. Calcium recovered when 4ml were used, pg-201, 199, 194. Calcium recovered when 6ml were used, pg-206, 208, 207. The volume required was not affected by raising the temperature of the acid to 60" C.MUREXIDE INDICATOR- The titration procedure, which is based on the work of Schwarzenbach, Biedermann and Bangerter,3 depends on the stoicheiometrical combination of EDTA with calcium ions to form a complex compound, the equivalence point. being indicated by the change in colour, from pale pink to purple, of a small quantity of added Murexide (ammonium purpurate). This colour change, which occurs only in alkaline solution, can be represented as- Calcium ion + purpurate + calcium purpurate complex. (purple) (pale pink) When the concentration of calcium ions in the solution is high, the reaction proceeds to the right and the dominant colour is pink, but in the absence of free calcium ions the reaction proceeds to the left and purple is dominant.As with acid - base titrations the amount of indicator used is a matter of personal choice. It should be sufficient to give a well-marked colour change, but not enough to upset the stoicheiometry of the main reactants, particular1:y in dilute solutions.Oct., 19521 CALCIUM I N PLANT MATERIAL 53 1 Phosphate-This is the only ion in plant material that interferes seriously with the titration. Calcium phosphate is precipitated on adding the sodium hydroxide, and a direct titration with EDTA results in a premature and fugitive end-point. Estimations on a plant- ash solution that contained 212 pg of calcium were made without the preliminary separation of phosphate ions, and the calcium found was 205 pg (mean of three discordant results). Magnesium-A precipitate with Murexide is produced both in the presence and in the absence of calcium, but the end-point is not affected.The volume of EDTA required to titrate a solution containing 20 pg of magnesium and 200 pg of calcium was the same as for 200 pg of calcium alone. Iron and manganese-In the absence of calcium, these ions do not affect the purple colour of the Murexide and do not produce a precipitate. However, in the presence of calcium the characteristic colour of the calcium - Murexide complex is modified somewhat, and may make recognition of the end-point more difficult. The volume of EDTA required to titrate a solution containing 200 pg of calcium, 12 pg of iron and 12 pg of manganese was the same as for 200 pg of calcium alone. EFFICIENCY OF PHOSPHATE SEPARATION- through resin columns and the amount of phosphate in each effluent determined.EFFECT OF SOME IONS ON THE TITRATION- Aliquot portions of a plant-ash solution of known phosphate content were percolated Phosphate content of each aliquot, 15-6 pg of phosphorus. Phosphate recovered (mean of 8), 15-5 pg of phosphorus. METHOD SPECIAL APPARATUS- Micro-burette-Of 0.5-ml capacity. Tubes-Select a piece of glass tubing, of 7-mm bore, and draw off one end to form a coarse jet; flair the other end, 70 mm from the constriction. Prepare eight or more tubes and mount them in a rack of any convenient design. A distance between holes of 2 inches allows 15-ml beakers to be placed below each column. A rubber ring round each tube prevents it from slipping too low in the rack.Teat pipettes-Select a piece of glass tubing, of 7-mm bore, and draw off one end to form a coarse jet 2 inches long. Cut the tube 3 inches from the constriction and flair slightly. One inch from the open end constrict the tube to a capillary by thickening the glass, and fit a teat. Graduate by sucking 1 ml into the pipette and making a scratch at the meniscus level. REAGENTS- Construct one teat pipette for each resin column. All reagents should be of recognised analytical purity. Di-sodium ethylenediamine tetra-acetate (EDTA)-Dissolve 0-86 g of sodium hydroxide in about 800ml of water, add 4.0g of the disodium salt of ethylenediamine tetra-acetic acid, and dilute to 1 litre. Murexide indicator-Grind together thoroughly 0.1 g of Murexide (ammonium purpurate) and 50 g of sodium chloride.The indicator gradually deteriorates and should be discarded after about 2 months. Standard calcium solution-Weigh accurately 100 mg of calcium carbonate, which has been previously dried for 3 hours at 105" C and cooled, and dissolve in about 2 ml of N hydrochloric acid. Transfer to a 100-ml graduated flask and dilute with water to the mark. 1 ml = 400 pg of calcium. Sodium hydroxide solution-Dissolve 4 g of sodium hydroxide in 100 ml of water. Hydrochloric acid, 5 N-Mix 250 ml of concentrated hydrochloric acid (spgr. 1.18) with an equal volume of water. Resin-Grind about 10 g of cation-exchange resin (Zeo-Karb 215) in a mortar, sieve and take the material that passes through the 60 B.S. and is retained by the 120 B.S. sieve, soak in 5 N hydrochloric acid overnight and wash several times with water by decantation.Pack a small tuft of cotton wool loosely into the lower constricted part of each of the special tubes to form a pad 1 to 2 mm thick, and then pour a slurry of prepared resin and Standardise immediately before use. PREPARATION OF THE RESIN COLUMN-532 MASON: THE DETERMINATION OF SMALL AMOUNTS OF Pol. 77 water into the tube until the bed is 25 mm deep. Lightly press another tuft of cotton wool on top of the resin bed and wash the column with distilled water until the pH of the effluent is not less than 4, as shown by a test paper. Regulate the speed of flow through the column by tamping with a glass rod until between 2.5 and 3 ml of water runs through in 5 minutes. When not in use store the columns with the tips standing in water to prevent them drying.Always wash with one bed-volume (2 ml) of water immediately before use. STANDARDISATION OF EDTA- Transfer exactly 0-5 ml of standard calcium solution (200 pg of calcium) into a small specimen tube, by means of a pipette, and add approximately 0.1 ml of sodium hydroxide TABLE I1 RECOVERY OF CALCIUM FROM STANDARD SOLUTIONS Taken, 10 50 500 1000 Found, Pg 10.04, 10.04, 10.04, 10.04 49.50. 49.50, 51.51, 50.1'7 495.0, 497.6, 501.1, 501.1 998.0, 990.0, 989.3, 1000.7 by means of a small teat pipette, and then 10 to 15 mg of indicator, measured on the blade of a micro-spatula. Titrate with EDTA until the colour changes from pink to a definite bluish-purple. In the presence of some other cations, e.g., magnesium, iron and manganese, the end-point is difficult to distinguish at first, but with practice it can be precisely noted.Repeat the standardisation until concordant results are obtained. Let 1 ml of EDTA solution = y pg of calcium. PROCEDURE- Prepare a solution of plant ash in dilute hydrochloric or nitric acid4 and accurately transfer a convenient small aliquot (say 0.1 ml) into a 15-ml beaker and add 5 ml of water. Transfer the solution to the column by means of a teat pipette; use a separate pipette for each sample. Wash the beaker and column with two separate 1-ml amounts of water and reject the effluent. Elute the calcium with 6 ml of 5 N hydrochloric acid, collect the eluate and TABLE I:[I COMPARISON OF EDTA AND OXALATE METHODS Sixteen samples of the same plant material were ashed and the calcium contents determined at different times by both methods EDTA method- Calcium content, pg-206, 206, 212, 206, 206, 206, 209, 210, 213, 211, 212, 212, 212, Mean, pg-209.8 Standard error, pg-2-84 Coefficient of variation, yo-l-35 Calcium content, pg-202, 208, 209, 207, 207, 205, 211, 208, 208, 208, 216, 210, 218, Mean, pg-210.1 Standard error, pg-4.66 Coefficient of variation, y0-2.22 213, 211, 212 Oxalate method- 212, 214, 219 evaporate to dryness on a water-bath.transfer to a specimen tube by means of a teat pipette. tion of EDTA solution. Dissolve ithe dry residue in about 1 ml of water and Proceed exactly as for the standardisa- Wash the resin column with water till free from acid; it is then ready for use again. Run a blank through the whole procedure concurrently with each batch of samples.If xml of EDTA is the titre of the sample and bml the titre of the blank, then the calcium content of the sample is (x - b)y pg. The purity of the reagents should be such that b is small compared with x .Oct., 19521 CALCIUM I N PLANT MATERIAL 533 RESULTS The sensitivity, accuracy and precision of the method can be assessed from the data given in Tables I1 and 111. Replicate determinations on a standard sample of plant material (apple leaves) are consistent, and there is satisfactory agreement with the values for calcium determined by the oxalate method.4 From Table I1 it is seen that 1Opg of calcium can be determined accurately and the sensitivity is considerably better than is attainable by the oxalate method. The ease and speed of analysis is also improved, for one person can comfortably complete 32 determinations in a day. The author wishes to express his thanks to Mrs. M. Souter who carried out the experi- mental work. REFERENCES 1. 2. 3. 4. Willson, A. E., Anal. Chew., 1950, 22, 1571. Mason, A. C., Ann. Rep. E. Mulling Res. Sta. for 1950, 1951, p. 122. Schwarzenbach, G., Biedermann, W., and Bangerter, F., Helv. Chim. Acta, 1946, 29, 811. Mason, A. C., Ann. Rep. E. A!laZZing Res. Sta. for 1949, 1950, p. 111. E.%sT MALLING RESEARCH STATION MAIDSTONE, KENT May, 195.2

ISSN:0003-2654    

DOI:10.1039/AN9527700529

出版商:RSC    

年代:1952

数据来源: RSC