MONK THE EXAMINATION OF FORTIFIED FOODS 83 The Examination of Fortified Foods BY H. E. MONK BSc. F.I.C. FOR the purpose of this review fortification of foodstuffs is taken to mean the addition of one or more specific nutrients to an article of food whose content in that respect is thereby raised above the normal level. Up to the present time the substances with which foodstuffs have been fortified fall into two classes minerals (including iron copper calcium and iodide) and vitamins. It is not proposed to discuss the first class those needing information on iron copper and calcium will no doubt turn to the monographs of McCance and his colleagues,48 while the one instance of iodide addition likely to be encountered is iodised table salt and in this connection the simple method described by D ~ n n 4 ~ may be recalled.On turning to the vitamins we are confronted with such examples as the addition of vitamins A and D to margarine of A and C to cocoa and chocolate of C to citrous products, and of B to flour and bread. It seems likely that we may expect this list to be extended considerably and that even if development is held up somewhat by the war it will occur soon afterwards. Approaching the subject from the other end one may take the known vitamins and select from them those which may conceivably be added to foods; in the light of present knowledge this list would include vitamin A aneurine riboflavin nicotinic acid ascorbic acid and vitamin D. All these have been or shortly will be used in this country or America and existing methods for their estimation will therefore be discussed.METHODS OF EsTIMATIoN.-These methods can be classed as biological microbiological, chemical and physical; before passing to a detailed consideration it may be profitable to emphasise some of the inherent virtues and defects of the different techniques. Pride of place must be given to the biological methods since this line of approach was used in all the early work leading to the isolation of individual vitamins. In the examination of foods it might be thought that the biological assay possesses the advantage that the reason for the addition of vitamins to foodstuffs is to secure certain biological results in those eating them so that a biological estimation is the most logical and likely to be the most exact.To take an example-it may be assumed that the need for vitamin A in the diet is to promote normal growth in children for the anti-infective protection of the epithelium (especially of the upper respiratory tract) at all ages and for the cure of night blindness. If the sample to be assayed is butter a physico-chemical measurement of the vitamin A will not assess the full potency in these respects since the human organism is able to utilise carotene for these purposes as well. If the growth rate of rats is used as a criterion this difficulty is automatically overcome. Similar pitfalls await the analyst with vitamin B,, in which he must not overlook the aneurine present as pyrophosphoric ester and wit 84 MONK THE EXAMINATION OF FORTIFIED FOODS vitamin C in the estimation of which he must take account of the dehydroascorbic acid, which also has antiscorbutic activity.The advantage of the biological approach is however less than appears on the surface. Certain questions suggest themselves at once. Even for the rate of growth are deductions based on rats sound when applied to man? It is believed for instance that pigs use p-carotene less efficiently than rats.51 How efficiently does the test animal utilise the vitamin in its diet? The answer to this question is quite complicated depending among other things on how completely the reserves of the vitamin have been exhausted in the animal prior to the experiment on the non-vitamin composition of the basal diet,50s56 on the presence in that diet of adequate amounts of vitamins other than the one to be assayed,54 on the age and sex of the animal on the conditions of housing and on the time of the year.Many of these difficulties may be overcome by careful experimental technique but their very existence serves to undermine the belief that because vitamins are added to foodstuffs for nutritional reasons biological assays are therefore the best. Two other difficulties with the biological method may be mentioned together-the question of specificity and the difficulty of assessing the criterion chosen as indicative of vitamin deficiency. For instance with vitamin A doubts have been expressed as to whether the increase in weight in rats can be taken as specific for the vitamin even under the most careful experimental conditions. It is however relatively easily and accurately measured.The xerophthalmia test on the other hand is probably specific but it is difficult to decide what precise stage represents a cure of the experimentally induced conditions. Similarly with vitamin B, for which the increase in weight of rats is less specific but more easily and probably more accurately measured than either cure of retracted neck in pigeons or of convulsions or brachycardia in rats (Coward52). For vitamin C the specific “tooth” method involves greater difficulties of assessment than does the increase in weight of guinea-pigs. With vitamin D however the position is rather different since the method depending on the ash-content of bones is a t once specific and capable of fairly accurate measurement (though it is very laborious).The line test and X-ray methods although, perhaps not so exact are still specific. For this vitamin the increase in weight method is discredited. Speaking generally therefore the biological assay depends upon the choice and measurement of a single criterion whereby the effect of feeding at different levels may be judged. With the possible exception of vitamin D this criterion may be one easily measured but of doubtful specificity or one specifically associated with the disease due to deficiency of the vitamin when difficulties in measurement are likely to be encountered. Other disadvantages of biological tests are well known-they are generally expensive and time consuming and the results are not very accurate. By their nature however, they do lend themselves to an accurate expression of their degree of inaccuracy and in this respect they are superior to all other methods.Physical and chemical methods may be considered together; they can nearly always be resolved into two main steps-an extraction whereby the vitamin is freed as completely as possible from the many and complex substances with which it is associated in the natural product and a chemical or physical operation by which it is estimated in the extract. The chief difficulties in the extraction stage will be evident; they are first to ensure com-plete extraction of the vitamin present and secondly to guard against conditions which may lead to the complete or partial destruction of the amount extracted. Hindrances to complete extraction may be merely mechanical as with tough and resistant material or connected with the combined forms in which the vitamin is present.For instance vitamin B is stated to exist in milk in four Vitamins as a class are rather sensitive-some to heat others to oxidation most to changes in $H and a t least one (riboflavin) to visible light so that considerable precautions are often necessary to avoid destruction during and after the extraction stage. In the actual estimation whether chemical or physical the chief need is for specificity. This is not surprising when the complicated chemical structure of most of these substances is borne in mind together with the fact that there are at the same time often present in natural products compounds of similar composition and properties. Existing met hods will, no doubt be improved as a result of further work and considerable progress has already been made.It is for instance remarkable how well the apparently unspecific oxidation-reduction titration has served for the estimation of ascorbic acid MONK THE EXAMINATION OF FORTIFIED FOODS a5 The chief advantages of processes of this type are relative speed and (under the best conditions) accuracy. The chemist also feels more satisfied with a result that can be ex-pressed in milligrams than with one given in terms of a biological response. This latter advantage is of course largely though not completely illusory. Finally such methods are generally less expensive than biological methods although for some of them comparatively costly apparatus is needed to secure the best results.A general reference may be made at this point to the latest method to be applied to the estimation of vitamins. Developments of the polarographic processes in this field are as yet in their infancy since results have hitherto been obtained only on comparatively pure solutions and no series has yet been reported of assays of natural products. If and when, the polarograph is successfully applied to such estimations it appears likely that its main use will be in factory control for which many estimations are required in the shortest time upon intermediates and products whose composition is fairly well known and constant. A description of ordinary polarographic technique has recently been given by Coates and Smart.63 (See also “Polarogvaphy,” by Kolthoff and Lingane reviewed in this number p.118.) The last line of approach that remains to be considered is the microbiological. The original discovery of Snell and Strong,36 that the organism “LactobaciZZus casei” could be used for the assay of riboflavin is now being followed by the publication of methods for the estimation of other members of the B-group based on other organisms. So far the bacteriological method has not been tested very widely in this country. The basal vitamin-free media for the growth of the test organism are usually complicated and not so easy t o make as the more ordinary bacteriological media but once this is overcome the method seems rapid simple and inexpensive. American reports on the riboflavin assay are very enthusiastic. Compared with the classical biological method it has three outstanding advantages all dependent on the change in the “experimental animal” from the rat or guinea pig to the bacterium:-(i) preparation of the “experimental animal’’ is greatly simplified (ii) since the numbers of experimental animals in one test are increased from say forty at the outside to millions idiosyncracy and abnormal response is ruled out (iii) instead of having to assess a complicated biological criterion of response the response is simply growth or non-growth of the organism.So far as the riboflavin assay is concerned pub-lished reports seem to indicate that the new method combines the advantages of both the chemical and biological methods with the disadvantages of neither. In passing to a detailed discussion of the estimation of each vitamin it may be said that for those vitamins likely to be encountered in fortified foodstuffs the difficult expensive and time-consuming biological assays can now be replaced by other methods usually physical or chemical of at least an equal and often of greater accuracy.This is true except for vitamin D for which at the present time there is no really dependable non-biological assay. In the application of non-biological methods particular care should be taken to ensure that the extraction process is quantitative and is conducted under conditions which do not harm the extracted vitamin and to guard in the estimation itself against inter-ference leading either to too high or too low results. This is done mainly by checking against other processes and by recovery of added vitamin.Where the importance of the test warrants it a final check will be afforded by biological assay. Figures for suggested human requirements of the vitamins are given by Sherman3 and by Bacharach and Drummond4; further details about the chemistry of the subject will also be found in Sherman’s book3 and in many other text-books of biochemistry and nutrition. VITAMIN A.-The formula originally proposed by Karrer Morf and Schoppl is now well established; the vitamin is formed in the animal kingdom by the breakdown of certain members of the carotenoid class of which the most important is 13-carotene. Symmetrical addition of two molecules of water at the central double bond and subsequent fission leads theoretically to the formation of two molecules of vitamin A from one of 13-carotene.Unlike the latter most carotenoids are unsymmetrical but any molecule of this kind having at one end of the aliphatic chain a 13-ionone residue attached to the two necessary isoprene residues in the chain may on similar hydrolysis give rise to one molecule of vitamin A. This is borne out by biological study; in addition to /3-carotene there are several other provitamins A of which the most important are a- and y-carotene and cryptoxanthin. These provitamins have as is to be expected only half the biological activity of the sym-metrical /I-carotene. A difficulty so far unexplained however is that from a compariso 86 MONK THE EXAMINATION OF FORTIFIED FOODS of the spectrographic and biological results it appears that JQ-carotene itself is only half as active in animal metabolism as it would be if it were broken down into two molecules of vitamin A.This question and the relationship between the carotenoids and the vitamin, has been discussed by Morton.2 In case of need the animal body readily utilises the pro-vitamins and analysts have to bear in mind that a biological claim to vitamin A activity may fairly be based on the presence of such substances as well as the vitamin itself. Units.-Units have been established to meet the needs of the biological investigator, and the chemist should not lose sight of the fact that the meaning of the term activity is a biological unit equal to that of a defined weight of a standard concentrate or pure substance. For vitamin A the International Unit is equal to the activity of 0.6 microgram of the standard p-carotene kept for the Health Organisation of the League of Nations by the National Institute for Medical Research Hampstead London.This is issued in the form of a solution in coconut oil containing 500 units (or 300pg.) per gram. Vitamin A is unique in that its activity is defined in terms of another substance. Hence any determina-tion by physical or chemical methods necessitates the use of a conversion factor if the results are to be expressed in biological units. Spectrographic Estimation.-The intense absorption of its solution in the near ultra-violet is the basis of the most widely used method for the estimation of vitamin A-a method which gives satisfactory results except with materials of low potency (of the order of 500 I.U./gram).In general a preliminary treatment of the sample by means of saponification is neces-sary except for concentrates of a potency of 20,000 I.U./gram or more. By this concen-tration of the vitamin the irrelevant absorption at 325mp due to glycerides etc. is either removed or relatively reduced. The alcoholic potash used for the saponification must be colourless and should be prepared from alcohol distilled over sodium or caustic soda. The time of saponification should not be unduly prolonged; for instance 10 to 15 minutes is recommended for margarines and 5 to 20 minutes for fish-liver oils. In the extraction of the unsaponifiable matter several points must receive attention. The ether should be freshly re-distilled and free from peroxides; for this purpose it may be kept over sodium prior to distillation.It is not easy to avoid emulsions especially during the washing of the ethereal extract. On no account should the first aqueous washing be conducted with vigorous shaking; some manage this stage with gentle shaking only and others5 by pouring a fairly large quantity of water through the ether in the separator. During the evaporation of the solvent the highly susceptible vitamin should be protected from oxidation by an atmosphere of oxygen-free nitrogen. The unsaponifiable matter after cooling in nitrogen is dissolved and the absorption a t 325mp is determined spectrographically. The solvents used are cyclohexane and less often absolute ethyl alcohol; some American workers favour a mixture of 10 per cent.of cyclohexane and 90 per cent. of ethyl alcohol. Chloroform or ether is to be avoided. Twyman and Allsopp6 have described methods for the purification of solvents. A descrip-tion of the methods and appliances of U. V. spectroscopy is given by Brode.' It is important that all the operations from sampling to the photographic exposure should be performed without delay preferably on the same day. Determination of the E tTm should be made over the range 250 to 400mp and plotted against wavelength. The absorption at 325mp is taken with or without correction to be due to vitamin A. Whether the amount of irrelevant absorption is material or not depends on (1) the vitamin A potency and (2) the nature of the non-vitamin material in the solution examined.Edisbury* gives a good account of this question with reference to margarines fish-liver oils and concentrates. As the vitamin potency decreases the percentage of irrelevant absorption increases and renders the spectrographic method unsuitable for products containing 200 I.U. or less per gram. The E ::m. at 325mp having been determined and if necessary corrected it remains to convert it into I.U. by means of a factor. British workers use the factor 1600 and American workers 2000 to convert E i:m. 325 mp values to I.U./gram. Discussion of the causes of this discrepancy will be found in the papers by Morton2 and Wilkie.5 Edisbury,* advocates a factor of 1600 for fish-liver products and of 1200 for refined whale products Detailed accounts of the application of the spectrographic method are given by Edisbury (loc.cit.) for margarine fish-liver oils and concentrates; by Garratt,g for malt extract and cod-liver oil; and by the B.P. Addendum for cod- and halibut-liver oils MONK THE EXAMINATION OF FORTIFIED FOODS 87 Antimony Trickdoride Test.-The transient blue colour given when chloroform solutions containing vitamin A and excess of antimony trichloride are mixed has been shown by spectrographic examination to be due to two absorption bands-one in the yellow and the other in the orange. With concentrates the maximum absorption is generally at 580 and 620mpJ and the same applies to unsaponifiable matter from natural products but with oils there is a shift towards the violet and the maxima approach 572 and 606mp.Usually the absorption in the 620mp region is about double that at 580mp. Abnormalities are met with in oils of which the absorption in the yellow may equal or even exceed that in the orange. In such oils the 606mp band is enhanced by ageing or by short-term treatment with oxidising agents (Heilbron Gillam and Mortonll). These facts largely account for the chequered career of the “blue value’’ methods. Opinion at the present time may perhaps be expressed in the following way if possible vitamin A should be estimated by means of the absorption at 325mp but where results obtained in this way are not reliable (as with low-potency materials) it should be possible by carefully controlled technique to obtain a fairly satisfactory estimate from the antimony trichloride reaction.Edisbury* has described the application of the antimony trichloride reaction to margarine etc. and from this and other sources the following general account of the method is compiled. For low-potency materials the test should always be conducted on the unsaponifiable matter and with fish oils and concentrates results obtained directly on the sample should not be accepted unless they have first been checked against the 325mp absorption. The antimony trichloride reagent-a saturated solution in chloroform-should be free from alcohol. It is often purchased ready-made and may then need treat-ment with a trace of water to remove iron. Even when stored in the dark it is not very stable; it must be kept in bottles painted on the outside not in dark glass bottles.If other means are not available, a tintometer of the Lovibond type may be used but it is more satisfactory to employ a visual spectrophotometer such as the Hilger-Nutting. Any photoelectric colorimeter which can be rapidly operated and is sufficiently selective to isolate a narrow wave-band should also be adequate. In any event accurate results can only be achieved if the colour is matched or the absorption measured over a narrow and particular range of intensity; hence after preliminary trials the solution of the sample or of its unsaponifiable fraction must be diluted with chloroform to the necessary degree. For tintometer readings this range is 4 to 5 Lovibond blue units and with the spectrometer the optical density (log Io/I) should be 2 f 0.2 per cent.When the required dilution is obtained a small accurately measured quantity of the solution is taken to it are added 10 volumes of the antimony trichloride reagent and the colour measurement is performed at once. If the Lovibond instrument is used yellow and neutral glasses will be needed as well as blue to obtain the match but in the calculation of the result they are to be neglected. It should be remembered that the Carr-Price “blue value” is the number of Lovibond blue units given by 0.04 gram of sample in 0.2 ml. of solvent and that the factor for converting such values to I.U. is 32 (but cf. Edisbury,8 p. 491). It may be more convenient for purposes of calculation to work the results out to “blue units per gram,” which on the above bases would need to be multiplied by 1.28 to give I.U.Workers having to rely on this method for a number of determinations should endeavour to obtain their own conversion factor by comparison with results derived by other approved means. If a visual spectrophotometer is available the absorption at 620 and also at 580mp is determined. It may also be necessary to allow the blue colour to fade and to redetermine the 620mp absorption subtracting the value found from the first determination. Results are expressed as E (on the original sample) and to express them as I.U. the equation E ik. 620mp x 510 = I.U. may be employed. It is well known that certain “inhibitors” (among them skatole and unsaturated gly-cerides) delay or prevent the development of the blue colour. Interference of this kind is generally overcome by saponification, Modifications of the antimony trichloride reagent have been proposed for which it is claimed particularly that a more permanent colour is obtained.The addition of guaiacol and warming at 55” C. for 14 minutes results in a change from the initial blue to a purplish-red (Rosenthal and Szilard12) and it may be that along such lines future improvements in the test will be achieved. Vitamin A, &.-It has been established that the place of vitamin A in salt-water The blue colour may be measured in several ways MONK THE EXAMINATION OF FORTIFIED FOODS 89 The complete estimation consists of three steps-extraction of the aneurine oxidation, and measurement of the fluorescence. In the extraction stage the two most important points to remember are that the vitamin is easily destroyed at a pH of 6 or above and that in many natural products it exists in part at least in the form of its pyrophosphoric ester, co-carboxylase.In some materials flour for instance the vitamin is almost if not wholly, uncombined and a simple acid extraction followed by centrifuging is sufficient to provide the solution ready for oxidation. Recommended details for the acid treatment vary con-siderably as to strength of acid time and temperature of extraction and the ratio of volume of acid to weight of sample. Henne~sey~~ suggests refluxing on a boiling water-bath with N/10 sulphuric acid in amounts not less than 15 nor more than 50ml. per g. of sample as a satisfactory and widely applicable method. Much stronger acid has however, been used.Kent- Jones Amos and Martin,24 for instance employ 83 ml. of dilute hydro-chloric acid (25 ml. of conc. acid per litre) to 20 g. of flour for 1 to 2 hours at a lukewarm temperature and quote Hay’s practice of treating 20 g. of flour with 15 ml. of conc. hydro-chloric acid and 25 ml. of water for 7 minutes in a beaker immersed in a boiling water-bath. In general it appears that with very finely divided materials or those that are readily broken down a short extraction is sufficient and that if desired fairly strong acid may be used whereas with more resistant material longer extraction with weaker acid is desirable. Usually it will be necessary to make a preliminary ethyl ether extraction to remove fat, when the amount of this constituent is appreciable.When any material amount of the vitamin is present in combination enzymic hydrolysis is necessary. Many enzymes25 have been used but the usual procedures is overnight incubation with pepsin in hydrochloric acid followed by 5 hours’ digestion with taka-diastase. Booth combines the two digestions by overnight incubation with 0.1 per cent. papain and 0.1 per cent. takadiastase in acetate buffer at pH 4-0. Houston and his co-workers have shown2s that in liquid milk there may be four forms of vitamin B, as this may be present in the unesterified state or as pyrophosphoric ester and either of these may exist free or bound to protein. To obtain the whole of the aneurine pepsin and takadiastase treatment is necessary but they find that except with late lactation milks peptic digestion may be omitted without serious error.The material resulting from acid or enzyme digestion is next centrifuged to obtain a clear extract for oxidation. Filtration is to be avoided since this involves preliminary treatment of the filter-paper,ZO which may adsorb the vitamin. It is of course necessary to have at this stage an extract which does not fluoresce. If the clear centrifuged liquid is already fluorescent it must be washed with isobutanol prior to oxidation. This results in an increase in volume which must be measured and taken into account in the calculation. In the oxidation of the aqueous extract care must be devoted to three points (i) the addition of the various reagents must be made in the same order and with the same timing to sample blank and standards; it is now generally held that the ferricyanide should be added before the alkali; (ii) a sufficient excess of ferricyanide must be present; (iii) rubber bungs which may yield fluorescent material to the isobutanol must be avoided and if a stoppered cylinder is used so must grease.The thiochrome is extracted with isobutanol, and its fluorescence is matched against standard aneurine solution similarly oxidised. Turbidity in the isobutanol extract is usually cleared by addition of a small quantity of ethyl alcohol. For measurement of the ultra-violet fluorescence some form of mercury vapour lamp is necessary together with a Woods glass filter. Practice is required for visual matching but with this accurate results may be obtained-indeed some workers prefer visual matching.For photoelectric measurement the essential equipment is a photocell screened by ap-propriate filters together with a galvanometer which should be capable of giving a scale deflection of about 50mm. per microgram of aneurine. Diazo Methods.-These processes derive from the original method proposed by Kinnersley and Peters,s7 and depend upon combination between the aneurine and diazotised aromatic amines. In the present stage of their development they appear to compare un-favourably with the thiochrome process for the estimation of aneurine in foodstuffs being at the same time more troublesome and less sensitive (the amount of aneurine in the final aliquot portion should be 0.5 to 2-5 microgram for the thiochrome and 20 to 100 micro-grams for the diazo methods.See for example Hennessey=.) Promising results have been obtained with synthetic concentrates27 For the takadiastase digestion the pH must be adjusted to 4.0 to 4.5 90 MONK THE EXAMINATION OF FORTIFIED FOODS Fermentation Method-This has not been much used in Britain although it has been fairly extensively employed in America particularly in the examination of yeast. It was developed by Schultz Atkin and Frey,28 from an observation that the fermentation of dextrose by ordinary baker’s yeast is hastened by the addition of aneurine and that up to 40 micrograms there is a proportionality between the amount of CO produced and the amount of aneurine added. Disadvantages of the method are rather specialised apparatus, and some doubts as to its specificity for vitamin B,.An attempt recently made to improve it in the latter respect29 depends upon ‘,‘sulphite cleavage.” Sulphur dioxide splits the aneurine molecule into two halves neither of which catalyses the yeast fermentation and the difference between the carbon dioxide produced before and after treatment with sodium sulphite at j5H 5-0 is claimed to give a true measure of the aneurine content of the sample. By a modification of the fermentation method it is claimed that both aneurine and co-carboxylase can be estimated ~eparately.~? Polarograj5h.-Preliminary observations are reported by Lingane and Davis5? RIBOFLAVIN VITAMIN B,.-Riboflavin lactoflavin ovoflavin etc. are now held to be one and the same substance for which the first is the most generally applicable name.I t is the predominant substance in the complex at one time called vitamin G (B,) in America. Fhorometric Estimation.-Aqueous or methyl alcohol solutions of riboflavin have a yellowish-green fluorescence. Attempts based on colorimetry such as the matching of the yellow tint against potassium dichromate have not so far been very S U C C ~ S S ~ U ~ ~ ~ but more promising results have been achieved by measurement of the fluorescence in the ultra-v i ~ l e t . ~ ~ . ~ ~ If reasonable precautions are taken in this the problem becomes one of obtaining a suitable extract. Here the main factors appear to be that riboflavin is sensitive to light, that it may exist either free or in combination with phosphoric acid or phosphoric acid and protein and that extraction processes may derive fluorescent material other than riboflavin from natural products such as for example black pepper extracts from which are fluorescent owing to ~ i p e r i n e .~ ~ General methods of extraction involve refluxing with acid 0.1 N hydrochloric 0.25 N sulphuric and 0.5 N acetic (in 70 per cent. methyl alcohol) acids have been used at the rate of 3 to 10 ml. per g. of sample. Autoclaving with water or 0.1 N hydrochloric acid is used in the bacteriological method and may in some instances yield extracts suitable for fluorometric assay. For the destruction of fluorescent substances other than riboflavin two methods have been proposed The first depends on oxidation with permanganate in acetic acid,33 and the second on reduction with sodium hydrosulphite and stannous chloride followed by re-oxidation to riboflavin by vigorous shaking with air.= Milk and products containing casein need special treatment and detailed processes have been worked out for them as well as for yeast (cj.30). The whole of these extraction processes should be conducted in the absence of short-wave light. Measurement of the fluorescence involves exactly the same considerations as in the thiochrome process for aneurine and may be visual or photoelectric. S h a ~ ~ ~ bases a method for liver extract upon extraction from acid solution with chloroform followed by aqueous extraction of the chloroform solution. The aqueous extract is rendered alkaline and irradiated under U.V. light whereby riboflavin is converted into lumiflavin and is estimated after acidifying and extraction into chloroform against standard riboflavin solutions similarly treated.BacteriologicaZ Method-This the pioneer microbiological method is due to Snell and Strong36 I t depends simply on the fact that riboflavin is essential to the growth of “Lactobacillus casei” (Lactobacillus helveticus Nat. Coll. Type Cultures No. 41 13) and in minimal doses the growth response of the organism is directly proportional to the concen-tration of riboflavin in the nutrient medium. The extract of the test material is added in varying amounts to a riboflavin-free medium and each tube is inoculated with one drop of a centrifuged 24 hours’ subculture of the organism. A series of tubes containing a standard range of riboflavin solutions is similarly treated.The tubes are incubated at 37” C. The responses to the known and unknown amounts of riboflavin are assessed and compared (i) by measuring the turbidity as an indication of bacterial growth for which purpose 24 hours’ incubation is sufficient (ii) by titrating the lactic acid produced by the organism and for this two to three days’ incubation of the inoculated tubes is desirable. The basal riboflavin-free medium is a little complicated; it contains glucose cystine, inorganic salts yeast supplement and NaOH-treated peptone. Both the yeast supplemen MONK THE EXAM~NATION OF FORTIFIED FOODS 91 and the peptone need special treatment to make them riboflavin-free; this is accom-plished by action with basic lead acetate on the yeast supplement and photolysis with visible light for the peptone.In the preparation of extracts for the test autoclaving with 100 times the weight of water for 15 minutes at 15 lbs./sq. in. or refluxing with 3 vols. of N/10 hydrochloric acid is generally used. In each instance solid matter is removed by centrifuging and the acid extract is of course neutraiised. Some milk and egg products have been assayed direct. Subsequent workers in America have reported favourably on this method.30.31 PoZarograph.-Lingane and Davis5' announce the application of the polarograph to the estimation of riboflavin but their work has been on solutions of the pure vitamin and has not yet been extended to extracts from natural products. NICOTINIC AcID.-It has now been established that the P.P.or anti-pellagra factor is nicotinic acid or its amide. So far as is known no foodstuffs in this country have yet been fortified by addition either of the acid or the amide although there is a proposal in the United States for such addition to flour. Chemical methods of estimation are mostly based on rupture of the ring after rendering the nitrogen quinquevalent. Cyanogen bromide has been used for this and the aromatic amines chosen for colour development have chiefly been aniline or 9-aminoacetophenone. A review of these methods has been given by Waisman and El~ehjem.~s Acid extraction of the sample is stated to be the best; the process as a whole works best with animal tissue ; vegetable materials frequently give high results. Two microbiological methods have recently been announced both of which apparently estimate nicotinamide as well as nicotinic acids9*G5; also preliminary work with the polaro-graph on nicotinic acid.57 PYRIDOXIN VITAMIN B (Adermin) .-Further work seems necessary before any of the methods proposed for estimating this vitamin4* can be accepted as satisfactory.ASCORBIC ACID VITAMIN C.-Vitamin C was the first among the vitamins for which the structural formula was determined and also the first to be synthesized. This and its comparatively simple constitution probably account for the fact that its estimation, particularly by chemical means is on a more satisfactory basis than that of any other vitamin. Indophenol Titration.-Credit for the application of the method of titration with the oxidation-reduction indicator 2 6-dichlorophenol indophenol is generally given t o Tillmans.41 It has on the whole proved reliable and has been used for estimating the ascorbic acid contents of many hundreds of natural products.In essentials the process simply consists in acid extraction followed by titration with the indicator. Ascorbic acid is as is well known readily oxidisable. The reaction proceeds first of all to dehydro-ascorbic acid and is reversible. I--I I co o=c I ll O * dehydroascorbic acid - H0.C Ji-l H.C - I I-ascorbic acid H.C -a I HO.CH I CH,OH I CH,OH Further irreversible oxidation causes a breakdown of the molecule. I t is the reversible oxidation which proceeds in the actual titration and it is therefore particularly important to see that it does not occur during the extraction stage as it may very easily since many natural products contain ascorbic acid oxidase (now believed to be a complex of copper and a protein).Protection against oxidation is generally afforded by maintaining an acid pH. Trichloroacetic acid has been very widely used but metaphosphoric acid is now held to give better protection against catalytic oxidation by copper and iron. Somewhat widely varying strengths of acids have been employed but with metaphosphoric acid a 3 per cent 92 MONK THE EXAMINATION OF FORTIFIED FOODS solution (prepared from glacial acid sticks) is common. Long extraction is not necessary, nor is it usual to apply heat during the course of the process. Resistant tissue may require thorough grinding with sand and the dilute acid.McHenry and his colleagues45 claim that in some vegetables part at least of the vitamin C exists in combination with protein and that acid hydrolysis ( e g . by boiling for 10 minutes with 0.5 N hydrochloric acid in an atmosphere of carbon dioxide) is necessary before the total ascorbic acid can be obtained, but this has not been entirelv confirmed by subsequent investigators. Recently workers in India55 have brought forward somewhat detailed evidence to show the existence of combined ascorbic acid in cabbage and certain native foods. Higher results were obtained by a modified method involving digestion with warm hydrogen sulphide trichloroacetic acid extraction and subsequent treatment with ascorbic acid oxidase than by the usual process. Extraction should be hastened as much as possible and filtration expedited or avoided by filtering through muslin or use of the centrifuge.Once the extract is obtained it is necessary to proceed with the titration without loss of time and this should be carried out on the acid solution the best pH range being 1.0 to 3.0 to which the solution may be adjusted by addition of sodium phosphate or sodium acetate. Titration is made with a dilute solution of the indicator prepared by dissolving a weighed quantity in warm water, filtering and making up to volume. The usual strength is in the neighbourhood of 0.05 per cent. Should the amount of ascorbic acid be small it is advisable to reduce the strength of the dye solution in order to avoid “poising” effects. I t is also essential to use freshly boiled and cooled distilled water in making the dye solution and for dilution etc.other-wise dissolved oxygen in the water will affect the result. In making the titration the aim is to get it done quickly and as far as possible to standardise conditions especially as to time. Each worker has his own procedure but it will be found satisfactory to fix an end-point in which the pink colour persists for 5 seconds and to complete the whole titration in 2 minutes. When working with materials the ascorbic acid content of which is unknown it may obviously be necessary to repeat the initial titration in order to obtain the result under such conditions. The necessity for the precautions described lies in inter-ference which may be encountered from other reducing substances particularly sulphydryl compounds.These have in general been found to reduce the dye more slowly than ascorbic acid and with the precautions described reliable results may be obtained. Coloured extracts too dark to be dealt with by the usual method may generally be successfully titrated over a layer of chloroform mixing being ensured by bubbling oxygen-free nitrogen or carbon dioxide. Photo-electric colorimetry has also been successfully applied to such extracts. With materials containing sulphur dioxide the preservative may obviously interfere with the titration. Two ways in which this may be avoided are described by Mapson.66 Either 20 per cent. acetone is added to the extract before titration or the SO is swept out by a current of nitrogen and exhaustion in vacuo.The standard dye solution does not keep and should be freshly prepared at least every other day. It is standardised against pure ascorbic acid; this is generally sufficiently pure, but it may be assayed by titration with standard iodine. Dehydroascorbic Acid-Since ascorbic acid is so readily converted into dehydroascorbic acid which itself has anti-scorbutic activity it is always necessary also to estimate this substance. This is very simply done by means of a duplicate titration. The first portion titrated in the ordinary way The second is saturated with hydrogen sulphide corked and allowed to stand overnight; the hydrogen sulphide is then completely re-moved by means of a stream of oxygen-free nitrogen or carbon dioxide and the solution is titrated as before.Any increase in the second titration represents dehydroascorbic acid. Interference with this test is discussed by King.@ With foodstuffs it does not seem likely that serious interference will be encountered. Iodimetric Titration.-Attempts have been made to estimate ascorbic acid particularly in orange and lemon juice by oxidation with standard iodine solution. Although accurate results have been claimed in these instances there is an obvious lack of specificity in the reaction which in practice suffers from the further disadvantage that the end-point is often indistinct. Recently Ballentinem has claimed that this latter drawback may be overcome for citrus juices by titration with potassium iodate and this method has also been success-fully applied to cocoa and chocolate.u Estimation as FurfuraZ.-Mention should be made of Roe’s46 method since it depends The colour change is visible in the chlorof~rm.~ MONK THE EXAMINATION OF FORTIFIED FOODS 93 on an entirely different principle and may on occasion be useful as a check but it is some-what complicated and there appears to be some risk of obtaining furfural from sources other than ascorbic acid in foodstuffs.Polarographic Estimation.-A preliminary note of a polarographic method has been published by Kodicek and WenigJ4' but difficulties in application to foodstuffs yet remain to be overcome. VITAMINS D.-It may be convenient to list the most prominent of the vitamins D:-D, molecular compound of lumisterol and calciferol. D, calcif erol; the vitamin produced by irradiation of ergosterol.D, isolated from fish-liver oil or derived from the irradiation of 7-dehydrocholesterol. D is a laboratory product and is not likely to be encountered; its absorption a t 265mp is nearly equal to that of calciferol.60 D has the same absorption spectrum as calciferoPl and the same antirachitic value for rats but it is considerably more effective for chicks than the ergosterol derivative. The structural formula of calciferol is :-D, formed by irradiation of 22-dihydroergosterol. Of these the first is only of historical interest. Me / I Me Me Me HO. ~ k \ ~ C a c / e ~ Ha H Vitamin D differs only in that it has the cholesterol side chain : /"" Me I -CH-CHa-CH,CH%-CH \ Me joined to C-atom No. 17. It is generally believed that the common form of vitamin D in animal products and also the one produced by the action of U.-V.light on the skin is D,. Present chemical and physical tests do not distinguish between D and D,. Antimony Trichloride Test.-With antimony trichloride in chloroform solution the pink colour given by vitamin D reaches a maximum in about half a minute and after about 5 minutes slowly fades. Brockmann and CherP2 were the first to suggest its use as a means of estimation but considerable difficulties have been encountered and several attempts have been made to improve on it. Nield and his co-workersa claim a three-fold increase in sensitivity by rigorous purifi-cation of the reagent the addition of 2 per cent. of acetyl chloride thereto and spectrophoto-metric measurement of the colours at 500 and 550mp.Milas and his colleagues63 have tried to account and allow for interference with the reaction which they ascribe chiefly to vitamin A and cholesterol. The vitamin A band (peak 600 to 620mp) fades more rapidly, while the cholesterol .colour (absorption peak at 455 to 484mp) develops more slowly than the colour due to vitamin D (absorption peak at 500 to 520mp). By making three readings in a colour analyser at 3 10 and 30 minutes they attempt to isolate these effects and to correct the 500mp absorption for them but results have been erratic. More satisfactory results on oils of potency 10,000 I.U. or more per g. were obtained by a rather elaborate process involving saponification separ ation of sterols by cooling in methyl alcohol to -10 to -15" C.and filtering refluxing for 1 hour with maleic anhydride in dioxane extraction with ether and application of the colour reaction to the residue. My thanks are due to friends and colleagues in particular to Mr. A. L. Bacharach, Dr. D. J. Kent-Jones and Dr. A. J. Amos who have freely given information on matters in which they are specialists. 94 MONK THE EXAMINATION OF FORTIFIED FOODS ADDENDUM ON INTERNATIONAL UNITS. 1. The International Unit has been defined as follows: for vitamin A,* the activity of 0-6pg. of pure 8-carotene; for vitamin B, the activity of 3pg. of pure aneurine chloride hydrochloride for vitamin C the activity of 50pg. of pure ascorbic acid; for vitamin D the activity of lpg. of a certain solution of crude irradiated for vitamin E the activity of l p g .of synthetic racemic tocopheryl acetate. (anhydrous) ; ergosterol;? There are no international units for riboflavin nicotinic acid pyridoxin etc. 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. 36. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 61. REFERENCES Karrer P. Morf R. and Schopp K. Helv. Chim. Ada 1931 41 1036 1431. Morton R. A. Chem. and Ind. 1940 59 301. Sherman. H. C “Chemistry of Food and Nutrition,” 6th Ed. 1941 New York. Bacharach A. L. and Drummond J. C. Chem. and Ind. 1940 59 37. Wilkie J. B. J. Assoc. Ofl.Agr. Chem. 1940 23 336. Twyman F. and Allsopp C. B. “The Practice of S@eclr@hotometry,” 2nd Ed. 1934. Brode W. R. “Chemical Spectroscopy,” 1939. Edisbury J. R. ANALYST 1940 65 484. Garratt D. C. ad. 1939 64 795. British Pharmacopoeia 1932 2nd Addendum p. 16. Heilbron I. M. Gillam A. E. and Morton R. A. Biochem. J. 1931 25 1352. Rosenthal E. and Szilard C. id. 1935 29 1039. Gillam A. E. et al. id. 1938 32 118 405. Gray E. Le B. J . Biol. Chem. 1939 131 317. Wald G. J. Gen. Physiol. 1939 22 391. Seaber W. M. ANALYST 1940 65 266. Grass Driers’ Association id. 1941 66 334. Gillam A. E. Biochem. J. 1934 28 79. Jansen B. C. P. Rec. trav. chim. 1936 55 1046. Wang Y. L.,’and Harris,-‘L. J. Biochem. J. 1939 33 1366. Hills G. M. id. 1966. Booth R. G. Chem.and Ind. 1940 59 181. Pyke M. J. SOC. Chem. Irzd. 1939 58 338~. Kent-Jones D. W. Amos A. J. and Martin W. personal communication. Hennessey D. J. Ind. Eng. Chem. Anal. Ed. 1941 13 216. Houston J. et al. J. Dairy Res. 1940 11 145. Lerrigo A. F. personal communication. Schultz A. S. Atkin L. and Frey C. N. J. Amer. Chem. Soc. 1937 59 1467 and 1938,60 1614. Schultz A. S . et al. id. 1941 63 632. Kemmerer A. R. J. Assoc. Ofl. Agr. Chem 1941 24 413. Emmett A. D. et al. Ind. Eng. Chem. Anal. Ed. 1941 13 219. Shaw G. E. and Hind H. G. Nature 1941 147 209. Van Eekelen M. and Emmerie A. Actu Brevia Neerland. Physiol. Phavm. Microbiol. 1936 5, Hodson A. Z. and Noms L. C. J. Biot. Chem. 1939 131 621. Shaw G. E. Pharm. J. 1939 143 222. Snell E. E. and Strong F. M. Ind.Eng. Chem. ArPal. Ed. 1939 11 346. Westenbrink H. G. H. Enzymologia 1940 8 97; Abstr. ANALYST 1941 66 26. Waisman H. A. and Elvehjem C. A. Ind. Eng. Chem. Anal. Ed. 1941 13 221. Snell E. E. and Wright L. D. J. Biol. Chem. 1941 139 676. Waisman H. A. and Elvehjem C. A. Ind. Eng. Chem. Anal. Ed. 1941 13 224. Tillmans J. et al. 2. Unters. Lebensm. 1928 56 272. King C. G. Ind. Eng. Chem. Anal. Ed. 1941 13 228. Ballentine R. id. 1941 13 89. McLellan B. G. unpublished work. Reedman E. J. and McHenry E. W. Biochem. J. 1938 32 85. Roe J. H. J. Biol. Chem. 1936 116 609. Kodicek E. and Wenig K. Nature 1938 142 35. McCance R. A. et al. Med. Res. Council Sfiecial Report Series No. 187 “The Chemistry of Flesh Foods and their Losses on Cooking,” 1933; id. No. 213 “The Nutritive Value of Fruits, Vegetables and Nuts”; id.No. 235 “The Chemical Composition of Foods.” Dunn J. T. ANALYST 1928 53 211. Tainsh P. and Wilkinson H. Chem. and Ind. 1939 58 1051. Braude R. et al. Biochem. J. 1941 35 693. Hilger, London. Wiley New York. 77; see also J . ASSOG. Off. Agr. Chem. 1940 23 346. ~ * Vitamin A itself is now believed to contain about 3 x 106 international units per gram. =f It is ROW generally accepted that pure calciferol contains 40 X IW international units per &am DISCUSSION ON THE FORTIFICATION OF FOODS 95 82. 53. 54. 66. Coward K. H. ggBiological Standavdisation of the Vitamins,” Ballikre Tindall & Cox London, Gates A. C. and Smart R. Chm. and Ind. 1941 60 778. El Sadr M. M. et al. Biochem. J. 1940 34 601.Pal J. C. and Guha B. C. J. Indian Chem. Soc. 1939 16 481; Abstr. ANALYST 1940,65 523. Sen Gupta P. N and Guha B. C. id. 496 549; Abstr. ANALYST 1940 65 523. Ghosh B., and Guha B. C. id. 505; Abstr. ANALYST 1940 65 523. 56. Coward K. H. and Kassner E. W. Biochem. J. 1940 34 638. 57. Lingane J. J. and Davis 0. L. J. Biod. Chm. 1941 137 667. 58. Olliver M. ANALYST 1938 63 2. 59. Kinnersley H. W. and Peters R. A. Biochem. J. 1934,28 667 and 1938 32 1616. 60. Windaus A. and Trautmann G. 2. flhys. Claem. 1937,247 185. 61. Windaus A. et al. id. 1936 241 100. 62. Brockmann H. and Chen Y . H. id. 1936 241 129. 63. Milas N. A. Heggie R. and Reynolds J. A, Ind. Eng. Chem. Anal. Ed. 1941 13 227. 84. Nield C. H. Russell W. C. and Zimmerli A J . Biol. Chem. 1940 136 73.65. Isbell H. et d. id. 1941 138 499. 66. Mapson L. W. Cbm. and Ind. 1941 60 802. 1938. COUNTY LABORATORY WORCESTER January 1942 DISCUSSION Dr. E. €3. HUGHES said that he considered that Mr. Bacharach in his introductory paper had very clearly and forcefully outlined our knowledge on the subject of the value and possibilities of carefully planned fortification of food. Speaking as a food chemist he felt that it should be left to the nutritional experts to advise what fortifying materials should be added to food and in what amount; they could be certain that manufacturers were very interested and would gladly co-operate. It would never do for manufacturers to add this or that to food as they pleased the advice on what to do should come from some very authoritative source and manufacturers would expect that on such an important and serious matter the instruction should come from the Government.He had understood Mr. Bacharach to say that he objected to the attitude of the person who opposed fortification because it was unnatural but there was something to be said where certain proof had not been given for the claim that the addition of chemical materials themselves might mean the missing out of some constituent accompanying the preparation in its natural source. Mr. J. L. SWEETEN said that Mr. Bacharach had mentioned iodised salt as an example of a fortified food which suffered no loss on storage. He believed that in fact losses up to 50 per cent. occurred; the cardboard containers appeared to absorb most of the iodine lost by the salt.He asked Mr. Bacharach if he had not understated the case for fortification by emphasising the point that it should not worsen such properties as flavour keeping quality etc. He personally thought that fortification might often improve foods indirectly; for example the use of fats of better quality than used heretofore was essential to prevent serious destruction of vitamins A and D in margarine. He asked whether this improvement in quality might not make margarine a more stable medium than butter for these vitamins. The addition of ascorbic acid to jam had been investigated by his firm but unfortunately the present practice of most jam manu-facturers was such that a very large proportion of both natural and added ascorbic acid would be destroyed before the jam reached the consumer.As his firm was one of the two manufacturing ascorbic acid in England they were very interested in such an outlet for it since provided that the destruction during manufacture could be prevented it was the most practical large-scale method of distribution. A t present, however a tablet appeared to be the most efficient way of conveying ascorbic acid from the manufacturer to the consumer. Miss M. OLLIVER said that the firm with which she was associated had during the past 12 months, manufactured many hundreds of tons of jams fortified with ascorbic acid for the British Red Cross Society for despatch to British prisoners of war. With the suitable precautions adopted the percentage reduction of ascorbic acid during manufacture was very small and the loss of vitamin on storing the finished product was negligible.Mr. Monk had mentioned in his paper that the 2:6-dichlorophenolindophenol method of Hams for estimating vitamin C did not include the biologically active dehydro-ascorbic acid or the so-called combined ascorbic acid. The results of work by Dr. Harris and Miss Olliver (shortly to be published) had shown that the quantity of dehydro-ascorbic acid normally present in fruits and vegetables was of little practical significance and no confirmation could be obtained for the existence of combined ascorbic acid. Mr. G. F. CLARKE said that the loss of vitamin D when using margarine for cooking was less than was to be expected. Vitamin D was not added to margarine supplied to bakers or confectioners nor to that sold for cooking.Owing to the shortage of eggs the vitamin D content of the margarine now being sold had been doubled. Dr. F. WOKES said that in considering the quantities of vitamins that should be added allowance must be made for the variation in requirements of different nutrients at different ages. A complete set of tables of these requirements for the most important nutrients was published last year in the Journal of the A mevicavr Medic& Association. For vitamin B the requirement ran parallel with the food intake so that the diet of children need not contain a higher percentage of the B vitamins than the diet of adults. Vitamin D requirements were higher for children. That of course raised the question of milk rationing and the supply of larger amounts to young children.Dr. F. BERGEL illustrating the addition of one vitamin out of a whole group referred to a recent paper (Pvoc. StaflMeetings Muyo CZi.nic 1941,16,433) in which it was reported that persons supplied wit 96 DISCUSSION ON THE FORTIFICATION OF FOODS the vitamin B group but with little B, showed after about 130 to 190 days symptoms of neurasthenia, anaemia etc. which onlyresponded to treatment when the vitamin B deficiencywas made good. Seemingly, vitamin & a t the present moment was among the most important from the medical and nutritional point of view. With regard to the stability of ascorbic acid Dr. Bergel said that an approximately 3 to 6 per cent. aqueous solution of that vitamin was relatively very stable to air. The more concentrated the solutions, the more stable they appeared to be.Referring to the thiochrome method of estimating vitamin B, he observed that it was relatively tricky, as the conversion of vitamin Bl into thiochrome was only semi-quantitative and might be complicated by such side reactions as e.g. the formation of the “aneurine disulphide” recently described by R. R. Williams (Chem. Abstracts 1941 35 3635) which was claimed still to possess 60 to 70 per cent. of the biological activity of aneurine. Dr. S. K. KON asked why it was that in certain countries e.g. the United States and Germany the addition of vitamin D to milk was largely successful whereas here it had made no headway. He had been rather surprised to hear the suggestion that vitamins A and D in butter were unlikely to be stable-he thought the butter would “go off” organoleptically before any change in the vitamins would take place.Mr. E. B. ANDERSON replied that he thought that the American people must be more susceptible to propaganda than the English. Moreover if vitamin D were added to milk it would be difficult to pass the cost on to the consumer. Dr. H. D. KAY wished to emphasise one point the need for very careful and independent control of the amount of vitamins actually present in the foods after fortification both to ensure that the stated amount was present when the food reached the consumer and to make sure that there was not undue excess. Some years ago when he had been working in Canada-shortly after it had been shown that irradiation with ultra-violet light increased the vitamin D potency of foodstuffs and of ergosterol-all types of food were being irradiated in Canada and U.S.A.or were having irradiated ergosterol added to them-foods such as milk bread biscuits fats of various kinds; also for what reason he did not know cigarettes! A t that time he was working on the effects of irradiated ergosterol in producing “pipe-stem” aortas and other toxic effects in rabbits and was naturally rather disturbed as to how far uncontrolled fortification of staple foodstuffs might be leading to similar results in man. From every aspect if fortification of foodstuffs was to be permitted it was necessary to insist on independent control which would in effect prevent either too much or too little vitamin or other adjuvant being added to the nation’s dietary.Dr. A. GREEN said there seemed to have been undue emphasis on the loss of added nutrients since such loss was also suffered by those of natural occurrence. A recent American paper had shown that during the fermentation of dough some of the aneurine added to the flour was converted into the phosphorylated compound which some might consider to be the more “natural” form. Dr. J. SWORD said that in the second paper Dr. Amos referred to the choice of calcium carbonate in preference to calcium lactate because this was not available giving the impression that otherwise calcium lactate would be preferable. It was generally admitted that calcium carbonate would be completely assimilated if added to white bread in the amounts under con-sideration but it was doubtful if more than a very partial assimilation would result if the calcium were added as calcium lactate powder.Dr. A. J. AMOS said that he entirely agreed with Mr. Bacharach’s remarks concerning the necessity for properly-trained and qualified chemists. Replying to Dr. Green he said that work in progress but not yet published proved he thought conclusively that some phosphorylation of added aneurin did occur during the making of bread. Replying to Dr. Sword he could not say whether added calcium carbonate would or would not be more readily assimilable than calcium lactate. Dr. K. H. COWARD said that Mr. Monk had stated the difficulties of biological tests rather mildly. One point of procedure reduced difficulties-the use of international standards. If one made a test on similar animals on an international standard of reference at the same time as one made a test on a substance, such differences as temperature basal diet and many other factors were ruled out.These international standards were originally impure materials but within the last few years pure vitamins had been sub-stituted for some of them. If anyone preferred he could use a preparation of vitamin B quite independently of the international standard. The method of comparison would of course be biological still but the value would be stated as equivalent to a given weight of material not as international units. Mr. Monk had paid one tribute to the biological test in saying that the inaccuracy could be assessed. As a matter of fact methods of estimating the accuracy of a test could be applied to the chemical and physical as well as to the biological determinations.Dr. D. W. KENT- JONES remarked that after listening to Dr. Cox and Mr. Monk it was obvious that the Public Analyst was going to be rather busy. There was a difference of opinion as to how the testing was to be done and in the Ministry of Food scheme it was necessary a t this stage to have experienced people to do this. He would emphasise the great difficulty of doing this test. Even a t the present time when in the Ministry of Food only a few were doing the tests and agreement was getting better it was by no means perfect and he would hate to think of a hundred Public Analysts doing tests without a set test. He would emphasise the importance of knowing the limits of error when doing these tests.It was no good laying down a method unless people knew the extent of one’s error. Theyfirnust know exactly where they were standing. Mr. A. L. BACHARACH said that he must on no account be understood to advocate the fortification of foods as an alternative now or at any future time to such measures as would make available to all mankind a diet adequate for optimal health. A11 that food fortification could do a t best was to prevent certain specific diseases known to be due to dietary deficiencies. But even for this limited result to be achieved, it was necessary to understand and apply the principles he had put before the meeting and for this a complete mobilisation of the necessary scientific and technical skill was essential. He thought that there was perhaps another reason. He quite agreed that later the Public Analyst might be able to do it. There must be set tests NOTES 97 Mr. D. M. FREELAND said that Mr. Bacharach had mentioned the attempts made by firms to add vitamins to foods and emphasised the necessity for legislation to standardise these attempts. He was in sympathy with the view put forward by Dr. Cox and thought that most food manufacturers in this country would also welcome legislation. There were many wholesome manufactured articles of food on the market which were being made before vitamins were discovered. Time advanced competition became keener and vitamins arrived. The problem confronting the manufacturer now was when did he cease to be a purveyor of pure food and when did he begin to purvey weak medicine? The manufacturer who ignored behests to add vitamins etc. to his products soon found that another concern had adopted them and so a form of competition was evolved which could be regarded as unfair. There should be restriction on the sale of such products. He thought that there was a proper fierd for concentrated foods and added vitamins and that they should be prescribed for those needing them