A 7.3'/ fl February 10, 1&12. U. S. DEPARTMENT OP AGRICULTURE, BUREAU OF CHEMISTRY— BULLETIN No. 147. H. W. WILEY, ciukk of Bureau. COAL-TAR COLORS USED IN FOOD PRODUCTS. BY BERXILYRD C. HESSE, Ph. D. Exp* ___ m 1 ^o«n U.S. DEPOSITORY Issued February 10, 1912. U. S. DEPARTMENT OF AGRICULTURE, BUREAU OF CHEMISTRY— BULLETIN No. 147. H. \Y. WILEY, Chief of Bureau. COAL-TAR COLORS USED IN FOOD PRODUCTS. BY BERNHARD C. HESSE, Ph. D. Expert, Bureau Chemistry. w \sinv. \-\mi \t im::\ 18U. ORGANIZATION OF BUREAU OF CHEMISTRY. H. TV. Wiley, Chemist and Chief of Bureau. F. L. Dunlap, Associate Chemist, and Acting Chief in absence of Chief W. D. Bigelow, Assistant Chief of Bureau. F. B. Linton, Chief Clerk. A. L. Pierce, Editor. A. E. Draper, Librarian. Division of Foods, W. D. Bigelow, Chief. Food-Inspection Laboratory, L. M. Tolman, Chief. Food Technology Laboratory, E. M. Chace, Chief and Assistant Chief of Division. Oil, Fat, and Wax Laboratory, H. S. Bailey, Chief Division of Drugs, L. F. Kebler, Chief. Drug-Inspection Laboratory, G. W. Hoover, Chief. Synthetic Products Laboratory, W. 0. Emery, Chief. Essential Oils Laboratory, under Chief of Division. Pharmacological Laboratory, Wm. Salant, Chief. Chief Food and Drug Inspector, W. G. Campbell. Miscellaneous Division, J. K. Haywood, Chief. Water Laboratory, W. W. Skinner, Chief. Cattle-Food and Grain Laboratory, G. L. Bidwell, Acting. Insecticide and Fungicide Laboratory, C. C. McDonnell, Chief. Trade Wastes Laboratory, under Chief of Division. Contracts Laboratory, P. H. Walker, Chief. Dairy Laboratory, G. E. Patrick, Chief. Food Research Laboratory, M. E. Pennington, Chief. Leather and Paper Laboratory, F. P. Veitch, Chief. Micro chemical Laboratory, B.J. Howard, Chief. Physical Chemistry Laboratory, C. S. Hudson, Chief. Sugar Laboratory, A. H. Bryan, Chief. Sections: Animal Physiological Chemistry, F. C. Weber, in Charge. Bacteriological Chemistry, G. W. Stiles, in Charge. Enological Chemistry, W. B. Alwood, in Charge. Nitrogen, T. C. Trescot, in Charge. Plant Physiological Chemistry, J. A. Le Clerc, Chief. Food and Drug Inspection Laboratories: Boston, B. II. Smith, Chief Buffalo, W. L. Dubois, Chief. Chicago, A. L. Winton, Chief. Cincinnati, B. R. Hart, Chief. Denver, R. S. Hiltner, Chief. Detroit, II. L. ScnuLz, Chief. Galveston, T. F. Pappe, Chief. Honolulu, Hawaii, Kansas City, Mo., F. W. Liepsner, Chirf. ille, R. W. Bjj i "M. ( 'hUf. New Orleans, W. J. McGee, Chief. N*\v York, It. E. Doolittlk, Chief. Omaha, S. H. Ross, Chief. Philadelphia, C. B. Brinton, Chief. Pittsburg, M. C, Albrbch, Chief. Portland, < hreg., A. I.. Knisei r, Chief. si. Louis, D, Bj Bisbbi, ( hief. Paul, A. s. Mm hi 1 1, Chief. d Francisco, K. A. Gould, ( hief. Ban Juan, Porto Rico, A. E. Taylor, Acting iimah, \Y. I . Ill hn ii, ( hitf. Seattle, II. M. Loo mis, Chief. 2 LETTER OF TRANSMITTAL. U. S. Department of Agriculture, Bureau of Chemistry, 'Washington, D. C, January 31, 1911. Sir: I have the honor to transmit for your approval a report by Bernhard C. Hesse, a color expert, containing both original chemical work done in the bureau since the passage of the food law on anilin dyes used for foods, and a valuable and extensive compilation of the lit (Mature of the subject, especially with reference to the harmfulness of coal-tar colors and their physiological effects. These data formed the basis of the opinions stated in Food Inspection Decisions 76, 77, and 106, and are presented in detail as of scientific and practical interest to all those concerned in the use of coal-tar colors in foods, whether as manufacturers, food officials, or consumers. 1 recommend the publication of this report as Bulletin No. 147 of the Bureau of Chemistry. Respect fully, II. W. Wiley, Chief. Hon. James Wilson, Secretary cf Agriculture, CONTEXTS. Page. Introduction 9 Purpose of the investigation 9 Number of colors permitted 11 Quality and efficiency of colors permitted 12 I. Identity of coal-tar colors used in food products in the United States in 1907 15 Collection of samples 15 Classification of samples submitted 17 Green Table numbers 17 Source 18 Patents 18 Shades of color 19 II. Purposes of food coloring 23 HI. Food color requirements 25 Adaptability for special purposes 25 Proportion of coal-tar color used 26 Suitability of shades of permitted colors and mixtures of panic 28 IV. Conformity of food color market, 1907, to recommendations of the National Confectioners' Association, 1899 30 V. Some legal enactments relative to the use of coal-tar dyes 35 List of thirteen foreign legal enactments 35 Summary of colors permitted by .these legal enactments 35 Colors said to be permitted under the German law of 1887 37 Definitenese and detail necessary to effect quality control 40 State laws prohibiting the use of colors in certain foods, 1909 41 VI. Recommendations by associations and individuals as to use of coal-tar dyes as food colors 42 Oazeneuve and Lupine 12 Society of Swiss Analytical < ihemista id Tschireh 43 Eayser 44 \W\ l it National Confectioners' Association 45 Bchacherl 45 Classification of recommendations in the literal ure 46 Conclusions 47 VII. Recommendations made by United States color industries and trades to the Department of Agriculture 17 Antagonistic to all added artificial color Concerning restrictions and requirements 48 Vm. Investigations, other than on animals, bearing on the harmfulness of coal-tar colors l'l-'fter ;,1 Winogradow Eeidenhain Other authors M 6 CONSENTS. Page. IX. Compilation under the Green Table numbers of all information avail- able as to the suitability of coal-tar colors for food 56 General statements 56 Classification of opinions in literature and in legal enactments showing condition of the market in 1907 62 Classification according to chemical composition and suitability 64 Physiological action of coal-tar dyes 67 Summary of symptoms 67 Experiments on dogs 67 Experiments on human beings 70 Experiments on small animals 71 General statements 71 Complete detailed statement of all combined data 74 Abbreviations of authorities cited 74 Tabulation by Green Table numbers of physiological and other data 75 Alphabetical index of trade names of coal-tar colors 148 X. Dosage and symptoms 153 Confectioner's list as a basis for a rule 153 Lehmann's rules 158 Santori's work as a guide to a rule 158 Young's rule 159 XL Oil-soluble or fat colors 159 XII. Rules and reasons for selecting the seven colors permitted by F. I. D. 76. . 161 Statement of rules 161 Analysis of three recommendations made to the Department of Agri- culture 162 Process of elimination 166 Reasons for adding Ponceau 3R 167 Quality, cleanliness, and efficiency ' 169 XHI. Lists of colors subsequently recommended by individuals and asso- ciations 169 Ernst 170 M uttelet's interpretation of the French law 171 Second International White Cross Congress 172 Beythien and Ilempel 173 Summary of three preceding recommendations 174 Unpublished recommendations of a manufacturer 177 Belial 177 Conclusions 178 XIV Chemical examination of the seven permitted colors, 1907 179 Need of chemical control 179 First met hods of analysis used 181 Moid ure 182 Chlorin aschlorids L82 Sulphated ash 182 Total sulphur 183 Gut /.cit lest 183 Heavy metals 183 Total insolubles 184 lit Imt extractive 184 Results of chemical examination, 1907 184 Detailed chemical data on each permitted color 184 Recalculation of analytical data on basis of coloring matter pres- ent 187 Market quality of the seven permitted colors 190 CONTENTS. 7 Page. XV. Guides in determining degree of purity and cleanliness 192 XVI. Analyses of certified lots of permitted colors, 1Q09-10 198 Tabulation of results 198 Comparison of analyses made in 1907 201 Conformity of analytical data with theoretical composition 202 Arsenic determinations on 86 batches 204 Suggested requirements for certified colors 205 XVTI. Methods of analysis used in testing colors for certification 210 Introduction 210 Xaphthol Yellow 8 211 Ponceau 3R 215 Orange I 217 Amaranth 220 Light Green S F Yellowish 221 Erythrosin 222 Indigo disulphoacid XVm. Addenda Additional examination of coal-tar dyes Supplementary list of trade names of coal-tar colors XIX. Index of authorities quoted 227 ILLUSTRATION Page. Fig. 1. Apparatus used in the determination of arsenic 213 Digitized by the Internet Archive in 2013 http://archive.org/details/tarfosuseOOhess COAL-TAR COLORS USED IN FOOD PRODUCTS. INTRODUCTION. PURPOSE OF THE INVESTIGATION. For the purposes of the investigation reported in the following pages, the legitimacy of the coloring of food and food products under certain conditions is regarded as established; the ethical and dietetic aspects of the question of food coloring are not here considered. The means at hand for coloring food products may be conveniently classified as vegetable, animal, mineral or inorganic, and synthetic or so-called coal-tar colors or dyes. Representatives of each of these have at one time or another all been used in the coloring of food, and the laws of various European and American States have, from time to time prohibited the use of certain specified members or all of each or some of the foregoing classes. It is therefore obvious that even for the legitimate purposes for which food can be colored, improper means are at command, and some of these, if not all, have been prohibited by law at some time or another. It is the function of the present work to determine what members of the synthetic or coal-car colors should be considered legitimate for coloring fond-. It is confidently believed that the material collected in the following pages point- clearly and solely to the following conclusions: 1. Coal-tar dyes should not be used indiscriminately in foods. 2. Onrj specified coal-tar dyes should be used in foods. 3. Only tested and certified dyes should he used in food-. The work here reported has Furnished the basis for Food [nspection Decisions Nbs. 76, 77, and 106, issued duly 13, L907, September -J."). i ( .i<>7. and March 25, L909, respectively, and the investigation it-elf was practically terminated January 1. 1910. The effect of these decisions ha- been to restrict the coal-tar colors permitted for use in food- to seven specified and enumerated colors, until Buch time as it -hall he -how n with reasonable conclusiveness that other Colors Bhould be added to BUch li-t : and further, all coal-tar colors permitted for use in food are to be of a degree of purity and cleanliness acceptable to the Department of Agriculture, and are to be -o certified. o 10 COAL-TAR COLORS USED IX FOOD PRODUCTS. In order to avoid any uncertainty as to the chemical composition of the enumerated colors, direct reference is made in Food Inspection Decision No. 7G to a standard work in which such chemical composi- tion is clearly and unequivocally set forth. The relevant parts of Food Inspection Decision Xo. 76 are as follows: The use of any dye, harmless or otherwise, to color or stain a food in a manner whereby damage or inferiority is concealed is specifically prohibited by law. The use in food for any purpose of any mineral dye or any coal-tar dye, except those coal-tar dyes hereinafter listed, will be grounds for prosecution. Pending further investigations now under way and the announcement thereof, the coal-tar dyes hereinafter named, made specifically for use in foods, and which bear a guaranty from the manufacturer that they are free from subsidiary products and represent the actual substance the name of which they bear, may be used in foods. In every case a certificate that the dye in question has been tested by competent experts and found to be free from harmful constituents must be filed with the Secretary of Agriculture and approved by him. The following coal-tar dyes which may be used in this manner are given numbers, the numbers preceding the names referring to the number of the dye in question as listed in A. G. Green's edition of the Schultz-Julius Systematic Survey of the Or- ganic Coloring Matters, published in 1904. The list is as follows: Red shades: 107. Amaranth. 56. Ponceau 3 R. 517. Erythrosin. Orange shade: 85. Orange I. Yellow shade: 4. Naphthol Yellow S. Green shade: 435. Light Green S F Yellowish. Blue shade: 692. Indigo disulfoacid. Each of these colors shall be free from any coloring matter other than the one specified and shall not contain any contamination due to imperfect or incomplete manufacture. The reasons, broadly considered, which led up to these food inspec- tion decisions are given in concise fashion in this introduction. Looking over the restrictions placed upon coal-tar colors by the lawmakers of the various countries it will be found that certain colors are in some instances specifically prohibited and in other instances that certain specific colors, or classes of colors, and only such, are permitted for the legitimate purposes of food coloring. Private organizations, such as the Swiss Society of Analytical Chemists and the National Confectioners' Association in the United States, have also made recommendations permitting specific colors only, and in addition specifically prohibiting others. Individual authors have likewise made similar recommendations. The control of the quality of llio food colors practiced on the part of those Gov- ernments which restrict the use of coal-tar colors to certain individ- uals, so far as any publications show, has not been very extensive. The action taken against the use of coal-tar colors for food-coloring purposes has ranged all the way from absolute prohibition of their Use for any purpose whatsoever to the practically unlimited use in Legitimate food coloring operations of all but two of such colors. INTRODUCTION. 11 Intermediate between these two extremes we find the prohibition of a greater number than two, or of all the members of this class except certain specified colors, and even here with the restriction that they shall be used only for certain legitimate food-coloring purposes. It would be desirable to have a number of coal-tar colors of estab- lished harmlessness specifically permitted, particularly if the number be sufficient to meet all the legitimate demands arising in the food- coloring art. To prohibit only specified coal-tar colors and, by implication, to permit all the rest of this class, would allow the unrestricted use of newly discovered colors, and all other coal-tar colors not examined as to their effect on health. A limited list of permitted coal-tar colors which would make the use of all coal-tar colors outside of the permitted list illegal would properly protect the health and could work no substantial hardship upon those engaged in food coloring. Any such hardship would be avoided by providing that if it is shown that none of the colors of the permitted list meets certain legitimate requirements and that coal-tar colors outside the permitted list are capable of satisfying this need and are in and of themselves harmless the permitted list can be expanded by the proper authorities to meet additional needs or growing require- ments without exposing the public health to any risk. NUMBER OF COLORS PERMITTED. It will be shown in the following pages that in the summer of 1907 there were on the market of the United State- 80 different chemical individuals, or so-called coal-tar colors, offered for the coloring: of food. It has been known since 1888 that it is unsafe to attempt to predict the harmfulness or the harmlessness of coal-tar colors by inference or analogy; therefore an ideally perfect permitted list should contain only such colors as have each been examined physiologically, separately, and specifically, and their harmlessness determined by actual test. Out of the 80 colors referred to 30 had not been exam- ined at all, so far as the literature shows, and therefore their harm- leasnesfl is certainly open to question; 20 had been examined physi- ologically! and the published accounts with respect to their harm- lessness or their harmfulness are in each case contradictory; on 8 none but adverse reports were to be found in literature. Leaving only 16 out of 80 colors on the market which had been established with more or less certainty a- harmless- that is, the users of these s Colors were deliberat ely taking chance- with the public health, since the harmful nature of those 8 had for a long time past been known to those conversant with such subjects; the use of the 26 doubtful colors is more defensible than the use of the s known to be harmful. Out of the 30 of whose action nothing was known it can not he -aid 12 COAL-TAR COLORS USED IX FOOD PRODUCTS. how many are or are not harmful, nor can the risk forced upon the public health be satisfactorily measured. This brief summary must suffice for the present as a justification for the restrictions of the permitted colors to 7 in number. The full reasons for each and every step will appear in their proper places in the pages following. QUALITY AND EFFICIENCY OF COLORS PERMITTED. An examination of 30 specimens representative of the 7 selected permitted colors on the United States market in the summer of 1907 disclosed such a condition of uncleanliness of product, or care- less or improper manufacture, and the use of such utterly inferior qualities of products for food coloring purposes, that control over the quality of the seven permitted colors seemed necessary. The results of the work in the making and maintaining of standards of quality for each of these seven colors also justify this control. That there was in 1907, and for a year or more later, a considerable diver- gence of opinion among chemists as to what should be the proper quality requirements for these colors is shown by the fact that out of 72 foundation certificates offered in accordance with Food Inspec- tion Decisions Nos. 76 and 77, 57 were rejected on their face because they did not comply with the standards of quality then in mind, or then shown to be commercially attainable. Much objection has been made by many of those whose certificates were rejected on the ground that the standards then in mind were unreasonable, unjusti- fiable, and nonattainable. The actual results, however, are that with very few exceptions the standards in mind early in the work have all been exceeded in practice; the 41,000 pounds (20.5 tons) of certified colors now in existence made in 97 batches, or an average of more than 420 pounds per batch, are, with the exception of perhaps one or two first batches, far cleaner than was expected when the 57 certificates above referred to were rejected. There has been no complaint against the permitted colors for want of efficiency or for the possession of unsuitable attributes, which 1ms been pressed or sustained with any such earnestness as would rea- sonably be expected if the defects complained of were as great as they were represented. Complaints have been made against (he yellow, when used in acidulated fruit sirups, on account of its pos- sessing a bitter taste; the proof of this, so far as any lias been offered, was for a long time not of a convincing nature, and it was two years after the first objection was raised before any concerted or positive action was taken by those interested. The yellow has also been criticized because it is not sufficiently fast to light ; although it was satisfactorily shown that another yellow was faster to light than the permitted yellow, no one has maintained that the yellow desired INTRODUCTION. 13 was actually preferable to the permitted yellow wholly and solely because of its superior fastness to light. The blue has been criticized because it is not of the proper shade to permit of its use in the bluing of sugar, but the substitute offered therefor has not been supported by its sponsors in a way to indicate that a defect of serious magnitude exists. On the grounds of suffi- ciency and of efficiency the fist of permitted colors selected appears to have been justified by the absence of any real or substantial com- plaint against them, on either or both of these grounds, during a period of more than three years. None of the seven permitted colors is patented; their manufacture and their purification are open to all, and none of the 80 colors on the market in the summer of 1907, with perhaps one exception, had been discovered since 1891; in other words, the advances in the coal- tar industry from 1891 to 1907 had added nothing to the colors serviceable to the art of food coloring. The list of colors permitted in Food Inspection Decision Xo. 76 embraces, therefore, a sufficient number of colors for all legitimate food-coloring purposes, the coloring of fats, oils, butter, etc., excepted, for which no suitable color had been examined and reported in the literature as being harmless and fit for use in foods; they can be made by any one; no one can have a monopoly in any one of them by virtue of patents; any competent maker can make all or any of them and purify them to the required degree of cleanliness. The standards growing out of the control exercised by the Department of Agriculture are such as to insure that the colors used for food-coloring purposes possess a proper degree of cleanliness and such a degree of cleanliness is commercially feasible and is a commercial reality. The policy adopted in this respect is therefore justified not only from the viewpoint of the history of the attempts on the part of various governments to control the quality of food colors, hut also by the results actually obtained by its adoption. This policy of restricting food colors to certain chemical individuals and demand- ing that those possess certain qualities is in complete harmony with the following suggestion made to the commission on rules and regu- lations under the food and drugs act, at its hearing in New York, in September, 1906. Any kind of a harmless color should !><• permitted, provided ii is not a color generally known n> he poisonous, ox generally found t" be poisonous, or one that may be almost impossible to !„■ produced without continuing son a- poison within itself when finished and ready for use. Coal-tar colors, as a class, should not be prohibited, but all those coal-tar colors generally found to 1m- p. [sonoUS, I r which are hard to produce without containing some poisonous propertief when ready for use, should be forbidden the i)ri\ [lege of being used, or offered for sale for use in food. Under the provisions of section 2, we have this to recommend to the commi that every person sailing or using a coal-tar color in food or drink, should be required 14 COAL-TAR COLORS USED IN FOOD PRODUCTS. to secure, either on his own account, or from the person from whom he buys such color, a certificate to the effect that the identical color used has been tested for poisonous ingredients, and is, to the knowledge of the chemist making the test, absolutely harm- less. The chemist should be required to be a competent physiological chemist, and must certify as above under oath. This would mean not that each package of color would have to be tested, but that every batch would have to be tested, and the certificate would then be held to relate to every batch. Such tests should be made in the United States and the chemist certifying should reside in and be a citizen of the United States. It should not be deemed sufficient to have any particular brand of coal-tar color tested once, and a blanket certificate given, covering the whole brand as long as it may be sold, but every ounce of coal-tar color put out by a color manufacturer should be shown by actual test to be harmless. While the exact mode of reaching the end in view is somewhat dif- ferent from the one suggested above, yet the fundamental object, that each batch of color used in foods shall be specifically tested, and that such colors shall be harmless, is attained with reasonable certainty; and although there are colors other than these seven which are undoubtedly equally as little objectionable, and while it is true that the present policy contemplates the permitted use of but seven specific colors, yet that policy, as before outlined, is sufficiently broad and elastic to enable the addition of a color to the permitted list, when it is shown that such color really fills a need, not properly satis- fied by one of the colors already permitted or some combination of these, and is in and of itself harmless. There can be no objection to the expansion of the list to such an extent as to include every harmless coal-tar color in existence; but the burden of proving such real need and harmlessness is very properly placed upon those who are seek- ing such expansion. As far back as 1892 the following statement was made on page IV of the Leffmann translation of Weyl's book on coal-tar colors, in con- nect ion with the various European legislative enactments: "It is certain that none of these plans is even approximately satisfactory and the problem will be even more difficult of solution in the United States; indeed, it seems to me to be unsolvable." In view of this opinion the results of the food inspection decisions as herein shown may properly be regarded as, at least, a step in the right direction tow ard I lie solution of this problem. This opinion Is further supported by C. A. Neufeld who, in review- ing Food [nspection Derision No. 76, says: "The idea of permitting only specific selected coloring matters for use in the production of articles of \'<>o(\, and of excluding all other colors from such uses, must, in the interest of control of articles of food, be regarded ai an extra- ordinarily happy one; a similar regulation is to be urgently recom- mended lor our own country." (Zts. Nalir. Oenussm., 1908, v. 15, p. W.) IDENTITY OF COAL-TAR COLORS, 1907. 15 IDENTITY OF COAL-TAR COLORS USED IN FOOD PRODUCTS IN THE UNITED STATES IN 1907. COLLECTION OF SAMPLES. The question " Which coal-tar colors shall be permitted for use in coloring such food products as are to be consumed within the United States?" can be answered, "All colors now in use, or to be used for that purpose, provided they are harmless and necessary as defined on page 14." This involves the further questions : 1. Which coal-tar colors of the 695 different chemical individuals now on the world's markets are actually used in the United States for that purpose ? 2. If restricted to such coal-tar colors as are now in use in the United States for this purpose, would this be likely to hamper or inter- fere with the invention of other coal-tar colors suitable for the coloring of food ? It would be physically impossible to go to every user of coal-tar colors in food products in the United States and obtain specimens of the coal-tar colors so employed; this would be impracticable not only because of the large number of such users, and their wide geographical distribution, but also because they often do not know what they arc using, and further because of a reluctance, undoubtedly to be encoun- tered among many, to disclose the nature of the products employed. This is rendered more than likely by the attitude of some of the makers of coal-tar colors, or their accredited agents, as will be shown Later. However, the sources of coal-tar colors are limited in number. By reference to panes IX and X of "A systematic survey of the organic coloring matters/' by Arthur G. Green, published in London and New York by Macmillan & Co. (Ltd.), in 1904 (hereinafter referred to as "Green Tables"), it will be seen that there are approximately 37 different concerns t be world over engaged in the manufacture of coal- tar colors. Therefore a canvass of these sources for such coal-tar colors as in their judgment, or in their business practice, they regard as proper for USC in I", is the best way of arriving at a fair demarcation of the held of coal-tar colors bere in question. Communication was therefore had with 13 actual manufacturers of coal-tar colors, in an endeavor to obtain from them BUch coal-tar colors as in their judgment or business practice are suitable loin are used in food products. A request was also made for information as to the chemical com- position of t he coal-tar col<>r specimens submitted j in order to avoid confusion, it was further asked that reference be made to the Green Tables, in which each chemical individual or coal-tar color ha- its 16 COAL-TAR COLORS USED IX FOOD PRODUCTS. own number, and if any of the contributed specimens was not so listed that the chemical composition be stated in a manner analogous to that used in the Green Tables. This procedure was necessary in order to reduce the terminology to a common and nonequivocal basis. Out of the 13 makers, or their accredited sole importers or selling agents in the United States, who were consulted, 9 have supplied the specimens requested; the remaining 4 promised to contribute, but have not done so. In the following table is shown the amount and character of the information obtained: Tabulation of distribution of replies and character of information received. Geographical distribution. Number of samples contri- buted. Number of samples referred to Green Tables. Number of samples not referred to Green Tables. No Country. Coal-tar color makers. Makers- asked to send samples. Makers not sending samples. Makers sending samples. composi- tion or ambigu- ous ter- minology. Germany 16 6 1 1 3 2 6 1 181 12 106 12 1 74 England 8 France 5 1 3 Switzerland... 5 1 2 51 35 5 21 Holland 1 Belgium Total 37 13 4 9 254 153 6 95 In order to make provision for the 24 makers listed in the Green Tables and not included in the 13 makers addressed requests for samples were sent to two domestic houses which import coal-tar colors from scources other than the above, for use in food products; their products must fairly represent any of the colors not covered by the 13 makers addressed. Of these two importers, one responded with 13 samples, and of each he gave the number in the Green Tables corresponding to each specimen; the other importer has not redeemed his promise to contribute specimens. A third importer volunteered the Green Table numbers of hair out of five coal-tar colors used in his business, but could not even approx- imately say what the remaining color was chemically. He did Q01 contribute any specimens, nor was that necessary ;it the time this informatioD was volunteered. A fourth importer contributed specimens of five coal-tar colors needed in his business, but was able to give Green Table numbers for only three of them: he could not give eves approximately the chem- ical composition of the remaining two. Out of the 17 responsible concerns consulted 5, or 29 percent, have not found it to their interest to contribute either specimens or information. IDENTITY OF COAL-TAR COLORS, 1907. 17 CLASSIFICATION OF SAMPLES SUBMITTED. GREEN TABLE NUMBERS. Out of the 284 specimens contributed, or reported on, 172 (60.6 per cent) were identified as to their chemical composition, by refer- ence to the Green Tables; 6 (2.1 per cent) were otherwise unequivo- cally identified chemically, and for 106 (37.3 per cent) the makers, or their responsible agents, declined to state the chemical composi- tion, i. e., 62.8 percent were unequivocally identified, and the remain- ing 37.3 per cent were not so identified. The specimens submitted are therefore divisible into the following three classes: Class I. Those for which numbers were given in the Green Tables, numbering 172. Class II. Those whose composition was given in chemical language, numbering 6. Class III. Those whose composition was not given in any lan- guage capable of correct and certain translation into chemical terms, numbering 106. Consider Class I. The Green Tables, page VI, divide the coal-tar colors into 21 groups, comprising 695 different chemical individual coal-tar colors. The 172 members of Class I number in all 74 indi- viduals, or 10.6 per cent of the Green Tables, and fall into 11 of the 21 groups of those Tables. The following table classifies the samples according to the Green Table groups: Green Table groups and number of collected samples falling within (hen}. Color groups of the Green Tables. Number of mem- bers in group. Collected samples falling into the several groups. Color groups of the Ben Tables. Number of mem- bers in group. Collected samples falling into the several groups. Number. Percent. Number. Percent Nitro 204 11 5 1- -' U l 30 11 l»i 24 5 Anthracene Indopbeool 37 38 32 9 ) 21 7 2 5 ■ Tctruk i-.:/.o 2 1 20 Btllbene 1 tone Sulphld.. Diphenj Imethane.. Triphciix Imethane. Xanthene 1 i:> 9 H Iildi^u 1 i:. 74 Acri.lm.... 97291 c r.ull. 147—12- 18 COAL-TAR COLORS USED IX FOOD PRODUCTS. SOURCE. The distribution of the 74 different chemical individuals of Class I among the 12 different sources from which they were obtained is as follows : Distribution of the 74 different samples of Class I among the 12 sources supplying same. Number. Percent. Number of sources from which each came. Number. Per cent. Number of sources from which each came. 35 20 i 4 5 3 47.3 27 5.4 5.4 8.1 2.7 1 2 3 4 5 6 1 1 1 1.35 1.35 7 8 9 10 11 12 1.35 It follows from this table that there is very little unanimity among the different concerns furnisliing coal-tar colors for use in food prod- ucts as to which of their products are desirable, necessary, or suitable for such use. Inspection of this table shows that only three colors out of 74, or 4 per cent, were wanted by more than half of all the sources; that only 6, or 8.1 per cent, were wanted by half of the sources; and that not one of the colors was wanted by all the sources. This last state- ment is true of manufacturers as well as importers, each group taken by itself. PATENTS. This lack of unanimity is not due to the patent situation, because not more than one of these 74 products is patented, and it is more than likely that the United States patent on this product has long since expired. Moreover, only 6 of the 12 sources offered colors at one time patented by themselves or others. The total number of such ex- patented products is 45, and of these only 22 were offered by those who had patented them; the remaining 23 were offered by sources other than the ex-pat cut ees, and were not offered by such ex-patentees. Patented colors. IDENTITY OF COAL-TAR COLORS, 1907. 19 It would, therefore, seem to be rather clear that others think more favorably of such ex-patented products as food colors than do the original patentees. In view of the fact that the latter would generally be in a better position, and would have greater opportunity than any one else to judge of the suitability of the patented products for use in food products, it may well be inferred that such products are not altogether free from disadvantages as food colors. The second of the two questions propounded, namely, If restricted to such coal-tar colors as are now in use in the United States for tins purpose, would this be likely to hamper or interfere with the invention of further coal-tar colors suitable for the coloring of food products? can be answered "No" because none of the colors submitted was discovered later than 1891; out of the 214 coal-tar colors since then discovered not one was among those submitted for use in foods, and out of the 481 discovered in 1891 and prior thereto, only 74 were so submitted, or 2 out of every 13 of such colors. In the following table these data are given year by year: Coal-tar colors discovered from 1740 to 1891. Year of dis- covery. Num- ber of sub- mitted colors discov- ered. Total num- ber of coal- tar colors discov- ered. Year of dis- covery. Num- ber of sub- mitted colors discov- ered. Total num- ber of coal- tar colors discov- ered. Year of dis- covery. Num- ber of sub- mitted colors discov- ered. Total num- ber of coal- tar colors discov- ered. Year of dis- covery. Num- ber of sub- mitted colors discov- ered. Total num- ber of coal- tar colors discov- ered. 1740 1856 1859 1861 1862 1863 1867 1 1 1 •-' 1 1 1 1 a i 3 5 3 3 Is71 \s~\ 1^7.5 is:»; 1877 1878 1879 1 6 5 ■J 12 7 r> 4 12 9 10 26 23 1881 1882 1883 1884 1885 1886 4 5 9 3 2 3 11 20 25 16 21 32 1887 1888 L89Q 1891 Total. 3 29 36 46 33 74 378 SHADES OF COLOR. The sufficiency of the 7 1 colors u>o 3 1 5 7 Q ID of ilycs offered fur cub ■had*. UN 106 1.^ 11', lOfl UN 1H7 u 20 COAL-TAR COLORS USED IN FOOD PRODUCTS. Green Table numbers of the 74 submitted colors showing shades and number of sources supplying each — Continued. Number of sources out of a possible 12 offering each color. Total number Shades. 1 2 3 4 5 6 7 8 9 10 of dyes offered for each shade. 502 518 520 523 512 448 504 517 7 54 516 53 64 56 89 269 329 9 Yellowish red 3 Scarlet 55 3 1 Yellow 510 8 94 4 7 Reddish yellow 84 14 17 85 95 650 2 13 2 667 18 97 439 476 480 655 440 468 464 287 433 434 425 2 Orange 86 6 Blue 692 6 Green blue , Violet 452 451 3 , Green 398 435 4 427 x 428 1 101 137 139 8 201 197 188 601 , BltW to bluish r«' i ■_* 1 iLiiki/.e.l Bfurei in ; psnnlttsd bj P.I. D. ■• 22 COAL-TAR COLORS USED IN FOOD PRODUCTS. Xine manufacturers sent 261 specimens, an average of 29 each, distributed as follows: 70; 38; 20; 20; 15; 43; 25; 18, and 12. Two importers sent 5 specimens each, and one 13, a total of 23 specimens, and an average of 8. These figures reflect a diversity of opinion as to what is needful for food coloring, since each one of these 12 makers or importers believed that for all practical food-coloring purposes his selection was complete and sufficient. Classifying the 284 specimens as red, yellow, brown, orange, blue, green, violet, and black, the following table is obtained showing the different requirements of each of the 12 makers or importers to produce the necessary shades of the eight colors mentioned: Total specimens submitted, grouped by makers and colors, showing number of shades required by each. Maker's number. Red. Yel- low. Brown. Orange. Blue. Green. Vio- let. Black. Number of shades wanted by each. Tot^.l speci- mens. 1 25 10 7 11 6 4 6 5 6 13 3 3 18 6 5 10 2 3 2 4 4 10 1 1 14 2 '"'io" 2 1 ..... 2 4 1 2 5 3 1 2 1 3 3 1 2 3 1 2 4 3 1 1 5' 5 1 1 2 2 1 2 2 1 3 2 1 i' ..... 7 7 7 7 7 7 5 6 6 7 3 3 70 25 18 43 20 15 12 13 20 38 5 5 2 3... 4 5 6 7 8... 1 4 4 1 2 1 9 10 11 12 1 Total.... Percent Maximum Minimum Average Permitted 99 34.86 25 3 8.25 3 66 23. 24 18 1 5.50 1 32 11.27 14 26 9.15 5 22 7.75 5 22 7.75 5 15 5.28 3 2 0.70 1 2S4 100 2.75 2.17 1 1.83 1.83 1 1.25 0.17 From this table it appears that not one of the 12 sources desired all of the 8 shades into which the 284 specimens are classifiable to make up a complete set of food colors; 7 out of the 12 sources wanted 7 of the 8 shades; 2 sources wanted 6 out of the 8 shades; 1 source wanted 5 of the 8 shades, and 2 sources were content with 3 out of the 8 shades. It will be noticed that the permitted list given in Food Inspection Decision No. 7(> provides for 7 dyes covering 5 out of the S shades of the above classification. The .*5 missing shades are brown, violet, and black; the shades provided are red, yellow, orange, and blue. It will also be noticed thai <>n 1 lie whole 6 ou1 of the 8 shades were not wanted by one <>r more of the L2 sources. The italicized shades are the ones nol provided for by the permitted lisl of Food Inspection Decision No. 76. Brown was not wanted by 5, nor orange by 1. blue by 3, green by 2, violet by .*). nor black by 10. PURPOSES OF POOD COLORING. 23 The combinations not wanted were as follows : Five sources omitted black only; 2, brown only; 1 blue and black; 1 brown and black; 1 green, violet, and black; and 1 brown, blue, green, violet, and black. Xot one of these 12 sources wanted only the three colors not found on the permitted list, and 2 sources did not want any of the three missing shades nor two of the permitted colors. In view of this large difference of opinion among the 12 sources as to the shades needed to make a complete set of food colors, the 5 shades selected for the permitted list of Food Inspection Decision No. 76 seem reasonably close to any consensus of opinion derivable from the tabulation of the collected facts. II. PURPOSES OF FOOD COLORING. The use of any color which conceals inferiority, or which gives an article an appearance better than it properly possesses is, of course, illegitimate, and such cases are not here considered. Among such uses may be mentioned that of color in pastry to impart a yellow color thereto, implying the presence of eggs, when they are either wholly absent or are not present in sufficient quantities to produce a shade of color which would indicate a superior quality. Such color- ing is frequently resorted to in macaroni, spaghetti, noodles, and the like, and it has also been stated in the literature that such coloring has the additional function of concealing dirt actually present in the flour. The addition of red coloring matter to meat products to give them an appearance of freshness which they do not of themselves poa the addition of red coloring matter t<> strawberry, raspberry, and similar jams, jellies, and preserves, to give them a color indicative of exceptional quality, even though they may contain none of the fruit \vh<»se presence is intimated by the label on the product ; the injection of red coloring matter into ordinary oranges to give them the appear- ance of blood oranges; the sprinkling of lemons and oranges with green coloring matter t<> give them the appearance of a particular origin or of a particular state of ripeness when such origin or state of ripeness ifl without foundation in fact; the injection of red coloring matter into watermelons to give them the appearance of ripeness, which ripeness they do not possess, are practices met more or less frequently. Among the purposes For which food colors are said to he used and the foods BO colored, the following are mentioned in the literature: In European emu,' 1. Macaroni ii colored \\ ith Dinitrocn and Martiui V- Hon . Handb i 2. Cordials and liqueurs with Dinitroci Arch. Pkarm., < 24 COAL-TAR COLOKS USED IN FOOD PRODUCTS. 3. Oranges: Biebrich Scarlet (163) (Weyl, Handbuch). 4. Pastry: Dinitrocresol (2) (Weyl, Handbuch). 5. Butter: Dinitrocresol (2) (Weyl, Handbuch). 6. To whiten flour: Anilin blue (457) (Zts. Nahr. Genussm., 1906, v. 12, p. 298). 7. Noodles are colored to cover up cigar butts, burnt matches, mineral oil, etc. (Zts. Nahr. Genussm., Vol. II, p. 1018). In the United States. 8. Jellies, fruit sirups, soda sirups, jams, ketchup, cheap cordials, lemon extract, milk, butter, cheese, ice cream, confectionery, pastries, flavoring extracts, mustard. cayenne pepper, sausage, noodles, wines, and liqueurs (Winton, Connecticut Agricul- tural Experiment Station Report, 1901, pp. 179-182). 9. Cattle feed is colored yellow (Gudeman, J. Amer. Chem. Soc, 1908, v. 30, p. 1623). 10. "Egg color" (399); ''Macaroni color" (94); "Tomato catsup color" (105); "Raspberry color" (103); "Mustard color" and "Pie filling color" (4); "Orange color" (87), and "Strawberry red color" (55) are corresponding United States com- mercial food color names and their corresponding Green Table numbers (Meyer, J. Amer. Chem. Soc, 1907, v. 29, p. 895). Dr. E. Ludwig, of Vienna, stated, upon the authority of Dr. Schacherl, at the International Congress of Medicine held in Budapest, in August, 1909, as follows: The rather widely distributed practice of coloring baker's goods yellow, such as cakes and the like, further the yellow coloring of pastry, macaroni, noodles, and so forth, has as its function the representation of a very large egg content in them; this coloring has been made very convenient because there are in commerce colors intended specifically for this purpose and designated "egg substitute" and which have nothing whatever in common with egg yolk. Marmalades such as apricot, raspberry, and currant marmalades are frequently found in a colored condition in commerce; in this case the purpose of the coloring is frequently to cover up adulteration; the adulteration may consist in an admixture of a cheap fruit pulp, particularly apple pulp, or in an addition of glucose sirup. Since these admixtures do not possess the color of the marmalades they are simply helped along by the aid of color. Old fruit sirups are toned up with color and then sold as fresh-6irup. Red colored fermentation vinegar and red colored vinegar essence as well as vinegar made from such essence are in commerce; such coloring has for its purpose to represent the product as "genuine red wine vinegar," which in sonic countries is highly desired. So-called "beer color," said to be an extract of roast <<1 mall (mall caramel), bin fact nothing but ordinary sugar caramel and is frequently from time to time publicly adver- tised; breweries themselves do not use this preparation, but it has been frequently shown that in small taverns by means of this color local beer was converted into Bavarian beer. Tin- wholesale coloring oi coffee beans serves the purpose of representing a better quality than it actually is. The coloring of cocoa and chocolate by tin' use of mineral additions and also of coal- often proven; in this case the coloring BGtVOI exclusively to cover up poor quality, in the case of good products such coloring is not practiced. I olored nusages, and in fact such with a colored meat body as will as such with a colored casing, are frequently colored ; coal-tar colors and cochineal serve this purj tin- latter, bowever, only for the meat This coloring is to preserve in old goods the ranee oi if' Tin- green canned goods of commerce are almost all colored w ith copper compounds. FOOD-COLOR REQUIREMENTS. 25 Tomato pulp frequently comes into commerce colored with a coal-tar color; the pur- pose of such coloring is to impart to the goods the appearance of having been prepared with extraordinary care. In all these cases it is not at all a question of a harmless change of the natural condition of the food product, but of improper manipulations which are adapted to deceive the purchaser as to the real value of the goods; even, indeed, to mask the danger to health. III. FOOD-COLOR REQUIREMENTS. ADAPTABILITY FOR SPECIAL PURPOSES. Not all coal-tar colors are adapted for use in food products. Colors are the more desirable for this purpose the higher their tinc- torial power, and the greater the resistance they offer to the action of the materials with which they are to be used, and under the con- ditions existing. Obviously only such colors as of themselves have their tinctorial properties fully developed can be used, and all such colors as require a mordant to develop or bring out the color are not fit for nor capable of use in food products. Further, if the colored material is subjected to varying tempera- tures in the process of manufacturing foods, it should be able to withstand the effects of such temperatures, as, for example, in the manufacture of candies. The colors should also withstand the action of reducing agents, such as are generated in the course of fermenta- tion and decomposition of the food product, or where a preservative such as sulphur dioxid is added to the food product to minimize the effect of decomposition of the food upon the color. Such colors arc put on the European market, and perhaps, but not necessarily, on the United States market with preservatives added to them. Most of the coal-tar colors are susceptible to the action of sulphur dioxid, particularly when the latter has been used in the decolor- izing of irhico^e, and Uranin (510) is one of the colors found to have the greatest resistance to the sulphur dioxid which may remain combined in candy. For example, the book entitled "Henley's Twentieth Century Book of Receipts, Formulas and IV ." published in 1907, on ;' sausage color: h i- absolutely necessary in using aniline colon t-« add a disinfectant t<> the dye- itufl solution, tin- <>l»jcj,. r,is 521; abst. '/J*. Nahr. OeMum^ I 16) report- a- follow- on the composition of colors used in coloring sausages, meat-, and presen es: l. Blood color: Moisture, r> per cent; common -alt, 6.6 per cent; boras, H per cent; and Ponceau 2B G 26 COAL-TAR COLORS USED IN FOOD PRODUCTS. 2. Blood red for meat juices: Liquid of a specific gravity of 1.0163 not affected by acids or alkalis and containing 27 per cent total solids; of these total solids 31 per cent were salt, 12 per cent borax, and the remainder Ponceau 2 R (G. T. 55). 3. Casing red: This powder contained Orange II (G. T. 86). 4. Sausage red: A liquid containing Eosin. 5. Lobster color: A liquid of specific gravity 1.0064 containing 1.64 per cent of solids, of which 10.9 per cent were salt and the remainder Ponceau R T (G. T. 44). G. Possetto (Zts. Ndhr. Unters. Hygiene }Yaarerik. 1891, v. 5, p. 105) cited the following 15 colors as being used for the coloring of pastry : G.T.No. G.T.No. Martius Yellow 3 Tropseolin 00 88 Tropseolin 000 No. 1 85 Tropseolin 000 No. 2 86 17 Chrysoidin . 18 41 Azoflavin 92 Victoria Yellow 2 Naphthol Yellow S 4 Aurantia Yellow 6 Acid Yellow G 8 Acid Yellow R 9 Citronin 91 Tropseolin 84 Algerian Saffron (a mixture of Nos. 4 and 86 and crocein). Prussian Saffron (composition not given). "Blood-red" on the American market is starch colored with red coal-tar colors {Ibid., 1896, v. 10, p. 114). "Butter yellow" is a clear saponifiable oil of reddish-yellow color, containing 3 per cent of anilin-azo-dimethylanilin (No. 16 of the Green Tables). PROPORTION OF COAL-TAR COLOR USED. The amount or proportion of coal-tar color used has been variously stated. On page IV of the Leffmann translation of Weyl's book entitled "The Sanitary Relations of the Coal-Tar Colors," it is stated that 1 ounce of Auramin (G. T. 425) will color 2,000 pounds of confectionery, which means 1 part of color in 32,000 parts of colored product. Frentzel (Zts. Nahr. Genussm., 1901, v. 4, pp* 968-974), on author- ity not given, says that for sirups 1 part of color is used to from 4,000 to 5,000 parts of sirup; in colored sugars 1 part of coloring inatter to from 1,333 to 4,000 parts of sugar; and in flour 1 part of coloring matter bo from 666 to 1,000 parts of flour. In pastry 1 to 100,000 parts (Zts. Nahr. Unters. Hygiene, }\aarenli. 1893, v. 7, p. 84). In chapter VII (p. 47), sections L5, 16, 17, and is, are brought together statements made before the commission on regulations for the Federal food and drugs act, as to the amount of color contained in colored food products. Briefly these ate as follows: Confec- tionery, l pari of color in 3,500 parts of product; beverages 1 part of color in 128,000 parts, 256,000 parts, 1,024,000 parts; butter, 420 grains of color to 1,000 pounds of butter; or 1 part of color to 16,666 parts of butter. FOOD-COLOR REQUIREMENT*. 27 One ounce of color to 30 pounds of "colored food;" the colored food was not further defined: which means 1 part of color in 480 parts of colored product. It has further been represented that 1 part of color is sufficient to whiten 250,000 parts of yellow sugar. From time to time others have presented information as to the amount of color used in food products. All of these data available have been tabulated, showing the number of parts of colored product containing 1 part of coal-tar color, arranged in the order of the amounts present : Food ! Do 2 Flour Do Sugar Confectionery. Sirups Sugar Sirups Confectionery. Butter Confectionery. 100 4S0 666 1,000 1,333 3,500 4,000 4,000 5,000 12,800 16, 666 20,000 Contectionery Do Do Be vera 24, 576 30. 000 B0, 000 Pastry 100.000 Beverages 120, 000 Do! 128 Confectionery LS Whitening sugar 250, 000 Beverages 256. 000 Do 1.024,000 Grouped according to the kind of material colored, the rai L r ivcn arc as follows: Be \< tragi « 80, 000 ; 120,000; 128.000; I O(M); l.HLM,000 l Biitt.-r b Confectionery 3,500; 12, 800; 20, 000; :»J;30,000; ' 32,000; 192,000 j Flour 1,000 Food ' 100; Pastry 100,000 Sugar 1,333; -1.000 Sirups i.ooi); 5,000 Whitening sugar 25 These statements have emanated From persons presumably ac- quainted with the facts of their own practice, and if that presumption rrecl it appears that there are wide variations in practice not only anions individual users, lull for individual colors. No attempt has been made to prove or disprove these statements l»y actual deter- mination <'f the amount of color contained in commercial colored food product-. The s <> chemical individuals on this market for food-Coloring pnr- . it can l»e fairly assumed, have been tested and tried out . their utility . and in this respect further tests were regarded a- super- fluous and therefore have not been undertaken. 1 Kind not > be for prtwrved tomatoes. 28 COAL-TAR COLORS USED IX FOOD PRODUCTS. SUITABILITY OF SHADES OF PERMITTED COLORS AND MIXTURES OF SAME. The shades produced by the seven permitted colors are, respec- tively, yellow, orange, blue, green, red, bluish scarlet, and brilliant cherry red. As statements are found in the literature against the use of all of the chemical individuals producing a brown or a violet shade, it will be necessary to produce these shades by a proper com- bination of two or more of the permitted colors. So far no criti- cism with regard to the shades produced by the seven colors them- selves and by their appropriate mixture has been made that has been substantiated. Objection has been made to the violet producible from blue ami red, on the ground that when applied to a food product, such as candy, the component parts do not evenly fix themselves upon the material. This objection, however, has not been pressed and proba- bly is not well taken, because of the fact, frequently reported, that few, if any, coal-tar colors are used without admixture of one or more other colors to shade or to tone the original color. Tins criticism, therefore, of the use of mixed colors can be regarded as not a serious objection. One criticism urged with considerable persistency against the seven permitted colors was that none of them would withstand the action of the organic acids ordinarily found in beverages such as lemonade, and it was suggested that no color was proper for use for such pur- poses which would not withstand, unaltered, for a period of 12 hours the action of a 10 per cent solution of citric acid. In urging tins objection substitutes were suggested for the permitted colors. The substitutes so urged were Tartrazin (94), Azorubin (103), Orange II (86), Ponceau 4 GB (13), and one other color designated as Scarlet SR, of whose chemical composition no information whatever was forthcoming. (The numbersin parentheses referto the Green Tables.) Of the five colors suggested it can be said that concerning all but No. 103 adverse statements are found in the literature, and X<>. 86 is spe- cifically regarded by every observer but one as being thoroughly poisonous. The suggested substitute list is, therefore, objectionable on the ground <>r Lnjuriousness to health. To best the validity of the assertion that, none of the permitted colors could withstand the action of citric acid, solutions of the sug- gested colors, as well as of the permitted colors, each one in a thou- sand, were submit ted to the action of citric acid, added in BUOh quan- tity that it amounted to io per cent of the total bulk of solution. This experiment showed that Tartrazin is reddened by citric acid, w hereas Naphtho] Yellow S loses in tinctorial power to a slight extent. The shade produced l>y Ponceau 4GB can bo closely imitated by a mixture of Naphtho] Yellow S, Orange J, and Amaranth, all per- FOOD-COLOR REQUIREMENTS. 29 mitted colors. There is no choice whatever in the shade produced by the desired Azorubin and the permitted Amaranth, nor is there any difference in behavior toward citric acid. The difference in the shade between the desired Orange II and the permitted Orange I is so small that it requires a side-by-side comparison to distinguish between them. Moreover, the desired Orange II produces a precipi- tate when brought in contact with the citric acid, whereas the per- mitted Orange I does not so precipitate. The permitted Erythrosin is, of course, completely precipitated by the citric acid. The per- mitted Light Green and Indigo disulphoacid are weakened in tinctorial power by the addition of the citric acid. Of these colors the only ones used to any extent in beverages, so far as either the suggested or permitted list is concerned, are red, yellow, green, and orange. As has been shown the permitted reds equal the desired reds and the permitted orange is better than the desired orange. The tinc- torial power of the permitted yellow is not so great as the tinctorial power of the desired yellow, but this difference is so slight that the objection urged against the list of permitted colors, namely that they were so poor in quality that they had destroyed a profitable and lucrative business in the coloring of beverages, is untenable in view of the fact that, assuming a price of $1 per pound for Tartrazin, and 40 cents for Naphthol Yellow S, and using them in the proportions necessary to produce a lemonade color in a 10 per cenl citric acid solution, it would take 5,000 quarts of finished lemonade to cause an increase of 1 cent in the cost of the production of the colored food product; that is, it. increases the price per quart by one five-thousandth of a cent. It has also been urged that the permitted green is not good enough for cordials and liqueurs, and that it is impossible to bring about the proper green by the use of the permit ted yellow and blue. This criti- cism, however, has not been persisted in; the fact is that mixtur the permitted yellow and blue can be made so as to obtain any desired shade of green, having a \ellow or blue Cast, and great clarity and brilliancy. How these mixed colors would look after a long period of time has not been a-cert allied. It has also been said that the permitted red, Amaranth, is not a color suitable for I he coloring of st raw berry jams and it has been urged that the same ehemieal individual under another eommereial name is better than the permitted red. This criticism has not been pressed, probably for the reason that, it can not he substantiated. \ am it was Claimed that the deposits to he noticed in bottled Lemonades Were due to Naphthol Yellow S, hut Solutions of Naphthol Yellow S in citric acid have remained without deposit for upward of i;> months; if is possible that such precipitation, if observed, may he due to an admixture of the nolipermit t ed Orange 11 with Naphthol 30 COAL-TAB COLORS USED IX FOOD PBODUCTS. Yellow S, and this difficulty can be obviated by the use of the per- mitted Orange I, which does not precipitate in the citric acid, as shown. Xaphthol Yellow S has been objected to on account of the bitter taste it is said to impart to the beverages to which it is added, and steps have finally been taken by those interested to have another yellow placed on the permitted list. It has also been objected that the permitted bhie is not suitable for the coloring of sugar, first, because it is soluble, and second, because of the unsatisfactory shade. It may be sufficient, in answer to this criticism, to state that there was no insoluble blue offered on this market, and there was no blue other than the one permitted offered against which adverse statements did not exist in the literature, and in view of this state of affairs the criticism may be said to be not well taken. As against all these specific criticisms it has been repeatedly stated by those in a position to know that they have found no difficulty whatever, by suitable mixtures of permitted colors, in reproducing any desired shade of any desired quality, not even excepting browns and violets. Considering all of these criticisms, therefore, the conclusion seems reasonable that there is no serious or permanent objection to be made against the seven colors selected, either as to qualities for food- coloring purposes or range of producible shades. IV. CONFORMITY OF FOOD-COLOR MARKET, 1907, TO RECOM- MENDATIONS OF THE NATIONAL CONFECTIONERS' ASSOCIA- TION, 1899. 1 Having thus shown that the food-color market of the United States contains not less than 80 coal-tar colors which are distinct chemical individuals, of which 74 are entered in the Green Tables and 6 are not, the next question to be considered is whether all of these sub- stances arc harmless and fit for use in food products. As a guide in determining this point the "Official circular from the executive committee of the National Confectioners' Association of the United St at es," pertaining to colors in confectionery, dal ed February 1 . 1899, may well be considered. The function of this circular is said to be "to throw light upon the vexed question of what colors may be safely used in confectionery," evidently because "there may at times be a doubt in the mind of the honest confectioner as t<> which colors, flavors, or ingredients he may safely use and which be may reject." The circular also states that "but infinitesimal amounts of color (coal-tar colors) need be or can be \\^^\ to give the desired effects," 1 See nlso p. 45. FOOD-COLOR REQUIREMENTS. 31 and in view of this statement as to quantity it must be self-evident that a color harmful when used in the small quantities said to be used in confectionery is certainly harmful when used in the large quantities used in coloring other food products. This circular under the head- ing, "Colors that are injurious and therefore to be rejected — Harm- ful organic colors/' enumerates 21 coal-tar colors. Of these 21 colors, 13, or 61.9 per cent, were among those submitted, and whose com- position was stated by reference to Green Table numbers; the Green Table numbers of these colors, together with the number of sources from which they were obtained, follow: Green Number Green Number Table of Table of numbers. sources, numbers. sources. 11 2 169 1 17 2 197 4 18 1 201 2 84 2 398 2 86 8 I 584 1 95 2 I 650 2 106 5 I Out of these 13 colors 3 each came from one source; 7 each came from two sources; 1 came from four sources; 1 came from five sources, and 1 came from eight sources, out of a possible 12; that is, one was wanted by more than half the sources. It is further to be noted that of the trade names given to the sub- mitted products of Class I the following appeared among the harmful list of this circular and also were found in identical form and spelling on the labels of the submitted products: 1. Bismarck Brown. 2. Chrysoidin R. 3. Chrysoidin Y. 1. Mandarin ( i extra. 5. Naphthol Green B. 6. Napthol Yellow. 7. New Coccine. 8. Orange A. 9. Orange A extra. 10. Orange G. 11. Orange II. 12. Scarlet. L3. Veeuvin B. 14. Croceio Scarlet 5 B. The following parallel will serve to show the great resemblance between the names given to the harmful colors of the circular and those found on the Bamples submitted: ( irruhir's }hu infnl I'tsl . Labi It af suhmittt l< s. Methylene Blue B B Methylene Blue B. Methylene Blue B B crystal* Methylene Blue D. Methylene Blue B BG Methylene BlueO Methylene Blue D B B New Coccine New I Z. \«\V ('..<( in. ■ z. Naphthol Green B Naphthol Green. 32 COAL-TAR COLORS USED IX FOOD PRODUCTS. Circular's harmful list. Labels of submitted samples. Orange II Orange II Z. Victoria Yellow Victoria Yellow Cone . Z . Victoria Yellow Cone. T Z. Acid Yellow Acid Yellow G. Bismarck Brown G Bismarck Brown B. Bismarck Brown T Bismarck Brown B 216. Bismarck Brown Dark. Bismarck Brown R X. Bismarck Brown T D. Bismarck Brown Y Bril. Bismarck Brown Y Dark. Bismarck Brown 2 R X. Chrysoin Chrysoin Brown G Z. Chrysoin G E Z. Chrysoin R E Z. Chrysoin R Z. Cochineal Red A Cochineal Red. Crocein Scarlet 3 B Crocein Scarlet 10 B. Crocein Scarlet 7 B Crocein Scarlet 8 B Fast Brown G Fast Brown N. Fast Brown 0. Fast Yellow Fast Yellow G. Fast Yellow 3 3. Fast Yellow 3 4. Fast Yellow Y. Imperial Scarlet, in powder, extra Imperial Scarlet 3 B. Safranin Safranin S P. Safranin A G extra. Safranin AGT extra. Safranin Cone. Safranin extra G. Safranin FF extra No. 0. Safranin G extra GGS. Safranin G 000. Safranin T. This comparison disclosed a considerable lack of conformity be- tween the United States food-color trade in 1907 and the circular of February, 1899, upon whose preparation for seven months prior to its date "a great deal of thought and labor have been given to a thorough investigation of the whole subject of 'colors in confec- tionery,' in which the committee lias been hugely aided by the researches of the association's chemist and by the results of his analytical tots" and whose "classifications have been carefully made, and are based upon the authority of the eminent chemists, Prof. Ivocuig ami Prof. Wcvl, upon the resolutions of the Swiss chemists and upon the French ordinances regarding the coloring of food products," and which list was expected to be ''of value to color FOOD-COLOR MARKET, 1907. 33 dealers and chemists/' and also was published to assist the confec- tioner in obeying "the letter and the spirit" of the pure-candy laws. Under the heading "Colors that have been shown to be harmless as used in the confectioner's art, harmless organic colors," this cir- cular enumerates 36 colors, for 4 of which there are no Green Table numbers. Of the 32 colors having Green Table numbers, 20, or 62.5 per cent, were among those colors submitted. The Green Table num- bers of these colors, together with the number of sources from which they were obtained, follow: Green Table numbers. 9. 13. 55. 65. 85. 103. 105. 107. Number of sources. ... 10 5 ... 1 ... 6 5 ... 2 ... 2 ... 6 ... 1 ... 7 Green Table numbers. 240 269. 287. 427. 448. 451. 462. 512. 517. 520. Number of sources. Of these 20 colors 5 each came from 1 source; 5 each came from 2 sources; 1 came from 3 sources; 1 came from 4 sources; 4 each came from 5 sources; 2 each came from 6 sources; 1 came from 7 sources, and 1 came from 10 sources; that is, only 2 were wanted by more than half the makers or importers, and only 4 by half the sources. It will be noted that this circular provides for a total of only 57 different coal-tar colors; the number of avowed colors submitted reached 80, or 23 in excess of this number, and further, that out of the 57 colors referred to in t his circular only 33 appeared among those colors submitted whose composition was acknowledged, ><» that for 47, or 58.8 ()<■!• cent of the avowed submitted colors, t his circular is no specific collide. From the data obtainable from this circular, the following tabulation can be made: Harmfukteu of tubmitted colon t>n list in circular. lot. 1. Total en market. market. Ciroilur's Branding list. n rata] . . Frrcciit. I'rrcent. These figures disclose a considerable and self-evident disregard of t he request . t hen eight years old, as made l»y t he National < 'on feet loners' 97291 ' Hull, l ff 12 — 3 34 COAL-TAR COLORS USED IN FOOD PRODUCTS. Association of the United States, that coal-tar colors designated by it as ' 'colors that are injurious and therefore to be rejected, harmful organic colors" be not used in confectionery, and by implica- tion that they should not be used in other food products. In view of this disregard with respect to 13 coal-tar colors out of 33 on the United States market in the summer of 1907, dealt with in this confectioners' list, the conclusion seemed justified that the action regarding other coal-tar colors on the United States market in the summer of 1907 was equally heedless or indifferent, a conclusion which is borne out by the material brought together under section IX. The necessity of adhering to some unequivocal terminology, as has been done in these pages by referring to the serial numbers in the Green Tables, appears from the following: Trade names are not definite with respect to the composition of the article sold under a given name, although tinctorially the colors may be substantially equivalent. In the Green Tables there are not less than 29 instances where the same trade name is applied to two or more different chemical individuals. In some cases these chemical individuals are fairly closely related, in others they are only remotely related. When the differences are greater than the presence or absence of a sulpho group or the use of methyl for ethyl or the reverse; the instances are given below: 1. Cotton Yellow 128 Primulin-azo-m-pheriylene-diamin-disulphonicacid. 191 Diphenylurea-riisazo-bi-salicylic acid. 2. Methyl Eosin 513 Methylated tetrabromo-fluorescein. 375 Dinitro dibromo fluorescein. 3. New Yellow 88 p-sulphanilic acid azo-diphenylamin. 91 Nitration product of diphenylamin yellow. 4. Orange ITT 23 Meta-nitranilin azo-R salt. 87 p-sulphanilic acid azo-dimethylanilin. 5. Orange X 43 Toluidin azo-Schaffer acid. 88 p-sulphanilic acid azo-diphenylamin. 6. Orange R 97 o-toluidin-monosulphoiiic acid azo-hetanaphthol. 15 Anilin-azo-R salt. 99 Xylidin sulphoacid-azo-hetanaphthol. 7. Toluylene Red 201 Dichloro-benzidin disazo R salt. 580 Dimethyl diamido toluphena/.in. The following quotation also bears on this point : In attempts i" group the aniline colors a kind of uncertainty appears even among color chemi -i . The same trade name docs not always correspond to the same prepara- tion. Many preparations are not chemical individuals, hut, mixtures of related colors. Many preparations an; 'standardized " for the trade; for example, with dextrin. On account, of the patent laws, factory secrets surround the produd LOO of many coloring matters, and frequently Statements are mot. with which are directly intended inv the purpose of mi leading competition, (Ilueppc, Die Methodcn dcr BaHcricnforschung, 5th cd., 1891, p. 105.) LEGAL ENACTMENTS. 35 V. SOME LEGAL ENACTMENTS RELATIVE TO THE USE OF COAL-TAR DYES. LIST OF THIRTEEN FOREIGN LEGAL ENACTMENTS. The 13 foreign legal enactments compared are as follows: 1. The Austrian regulation of March 1, 1886, which forbids Xo. 483 of the Green Tables and all anilin colors. 2. The law of Austria of May 1, 1886, which forbids Xo. 1 of the Green Tables and all anilin colors. 3. The Austrian regulation of September 19, 1895, in which cer- tain colors only are permitted, and all others are forbidden. There are 16 titles of permitted colors in this law, but these are in some cases so elastic and so indefinite as to include 47 entries in the Green Tables. 4. The Austrian law of January 22, 1896, in which there are 17 titles, but these are sufficiently elastic to allow of 22 entries in the Green Tables being included in them. 5. The German law of July 5, 1887, which specifically prohibits only Xos. 1 and 483 of the Green Tables. The interpretation which seems to be generally placed upon this law is that all other Green Table members are permitted in Germany for use in foods. 6. The Italian law of February 7, 1892, forbidding all colors except 9 different titles, which, however, were elastic enough to include 32 entries in the Green Tables. 7. The law of Italy of February 7, 1902, which prohibits .17 entries in the Green Tables, and permits 11 specifically. 8. The Italian decree of June 29, 1893, in which there were 7 titles of permitted colors sufficiently elastic to include 34 different individuals. 9. The Italian decree of March 24, 1895, forbidding four titles cov- ering only four entries in the Green Tables. 10. The French police ordinance of May 21, 1885, in which is 1 ' entries in the Green Tables were prohibited. 11. The French police ordinance of December 31, 1890, in which 469 entries of the Green Tables were prohibited, and which also permitted under 9 titles 23 entries in the Green Tables. 12. The Belgian law of 1891, which specifically forbids only four entries in the Green Tables. I."!. The law of (he Canton of Te»in, dated May Is, iso;. which forbids only -1 specific entries in the Green Tables. SUMMARY OF COLORS PERMITTED BY THESE LEGAL ENACTMENTS. An examination of the 13 legal enactments made in Knrope with respect to the use of coal-tar colors in food products discloses consid- erable difference of opinion as to the barmfulness or the harmlessnesfl of even the same chemical individuals. To prepare an approximate 36 COAL-TAR COLORS USED IN FOOD PRODUCTS. and comprehensive summary of the effect of such legislation, the following plan has been followed : The provisions of a selected number of laws and regulations, 13 in number and dating from 1883 to 1902, were read with respect to permission or prohibition of the 695 chemical individuals listed in the Green Tables. The effect of each law upon each separate entry in the Green Table numbers was noted, either as permitted, forbidden, or noncommittal when the law was silent upon such entry; the laws were read with the understanding that what was not forbidden was permitted, and what was neither forbidden nor permitted was non- committal; if this were an incorrect or improper procedure the number of permitted colors would be reduced but not augmented. For the purposes of this first approximation, no attempt was made to segregate those specifically permitted from those permitted by blanket expressions or phrases. For the purposes of a side-by-side comparison, the term " index number 1 ' was coined; this ''index number" gives in the first place the number of legal enactments that permit the color; in the second place the number of enactments that forbid it; and in the third place the number of enactments that are silent or noncommittal. Thus: 2S3 as an index number would mean 2 enactments permit, 8 forbid, and 3 are noncommittal; so that an index number with the highest hundreds would have the greatest number of permissions, and with the highest tens would have the highest number of pro- hibitions. With this understanding of these terms, the following table of so-called index or "P. F. N." numbers is offered: Thirteen legal enactments classified by Green Table numbers and the "P. F. N. figure," or "index number." Total number of Green Table entiles. Index number. Qreen Table numbers. l 1 3 217 1 366 29 3 1 7 ■J 2 3 3 11 1 ti 2 184 J74 2,10,1 304 373 891 468 IM 544 663 :»7i 643 661 (.70 71-' 751 081 483. 1-3. 394-411, 416-426, 428-434. |S5», 436-446, 484-492, 528-661, 564-683, 688 600.602-649,651-691,693-696. 108. 0,7, 10,11, 12,14, 10,19-22,35-40,42, 4:v«-V:>4,. r ,S-04. 66-83, 89, 90. 88, 84, 86- 101,104-106, L09-113, 115 L45, 149, L51-156. 168, L69, 161, 162, 164-168, 17_\ 197. 1'' • 1,412-414,493-611, (46, 1 1 178, 179, 181, 154,477,51 1,519,622. 662,601. 515,516. 480,514 44,11 i l ', L5 35,92,108,146 1 K 160, 160, 117. l, L8 ii. 169,512,518. 451, 162. 4, 55. 54,57,102 lis 467. 167,171. Italicized Agorei Indioate colors permitted by T. T. i>. 76, LEGAL ENACTMENTS. 37 From the table it appears that there is no one Green Table number that is permitted by each and all of these 13 legal enactments, nor is there any one color that is prohibited by each and all of these 13 legal enactments; and as late as 1902 there was considerable con fusion as to what should or should not be permitted or forbidden. That the European enactments were not in reality consistent or effective appears from the following: 1. In respect to the use of coal-tar colors, the views as to their harmfulness or hann- lessness are very divergent, and this uncertainty is expressed in the various legislative enactments, (v. Raumer, Zts. Nahr. Unters. Hygiene d'Waarenk., 1895, v. 9, p. 207.) 2. After eating groats, which no doubt were free from ordinary poisons but had been colored with Martius Yellow, a whole family became sick. Since this coloring matter is not mentioned among those which, under the law of July 5, 1S87, are forbidden for the purpose of coloring articles of food, a complaint could not be lodged. Never- theless, the use of Martius Yellow for the coloring of articles of food would seem to be dangerous, since this coloring matter exerts poisonous • It is a weakness in the law that coal-tar colors, of which new members arc continuously appearing on the market, and whose physiological action is unknown, should be at all permitted for the coloring of articles of food. (Dietrich, Th., Jahresberichte der landwirthschoft- lichenVersuchsstation, Marburg, 1900-1901, p. 13; abaft. Zls. Nahr. Genustm., 190£, r. 5, p. 364.) On account of their large number and the great diversity of opinion as to harmfulness of some coal-tar colors and the harmlessness of others therein reflected, no attempt was made to collect all the legal enactments and regulations made with respect to coal-tar dyes. The foregoing are typical of the remainder. COLORS SAID TO BE PERMITTED UNDER THE GERMAN LAW OF 1887. That some of the laws did not employ specific terms but used those possessed of a «;reat degree of elasticity appears from the following t aken from WeyPs "Sanfl ary relat i<>ns of \ lie coal-tar colors," page 38, concerning the Chamber of Commerce and Trade of Sonneberg which declared on December 4, 1887, that the German law of July .". L887, allowed the unrestricted use of — All blue an not less than L'.;:; permitted colors, as will now be shown. Using the Green Tables as a guide it will be found that under the above ruling there are to-daj no less than 107 blue coal-tar colors which could be wm><\ for f<»«..l coloring. The} are arranged as follows, showing the comments < he justified that the fact that a color has been permitted or has been forbidden by any one or the majority of t hoe Legal enactments ought n.»t t<» constitute a clean bill of health, n<»r ;i u indict men t . a-- the CaS6 may he. This side-by->ide comparison must ii"t he pushed t«- extremes; indeed the extent to which it can he employed LS naturally \«-ry lim- ited, and the purpose for which it was made was t<> reflect in a manner easily and comprehensively grasped the confusion and inconsistencies which very persistently force themselves upon the mind of a person reading those enactments and haying in mind the chemical individ- uals at which they are aimed. 40 COAL-TAR COLORS USED IN FOOD PRODUCTS. The definite lesson to be learned from this side-by-side comparison is that these enactments in many cases employed terms so vague and indefinite as to permit the use of some bad colors as well as all good ones, that is they were not sufficiently definite to exclude all that were harmful. DEFINITENESS AND DETAIL NECESSARY TO EFFECT QUALITY CONTROL. Tins apparent state of confusion in legal enactments that pre- ceded the summer of 1907 was a very strong factor in the formation of the conclusion that in order to be effective any law or regulation dealing with coal-tar colors for use in foods must prohibit every coal- tar color except certain definite specific ones. The Austrian laws of September 19, 1895, and of January 22, 1896, provided for quality control by public and other laboratories of the coal-tar colors put upon the market for use in foods; the results of such control, as reported in the Zeitschrift fur Nahrungsmittel- Untersuchung, Hygiene und Waarenkunde, 1896, v. 10, p. 335, are as follows: Coloring matters of commerce are mostly mixtures of various coloring matters, a right which manufacturers will not part with; and further, while it is indeed possible to test the coloring matter in substance, it is nevertheless impossible to test it in the very small amounts which are used in the coloring of foodstuffs and to determine with certainty the identity of the color as to whether it is or not one of the permitted colors. Of 21 samples of coloring matter examined, 14 were objectionable, partly because of false labeling, or because they were mixtures, partly because they contained poisonous metals, or a forbidden coloring matter. Thus, a so-called "Ever- green" was Naphthol Green B, a poisonous nitroso color; Malachite Green contained zinc; an Acid Magenta and a Rosalin contained traces of copper; Ponceau, Eosin, Brown, and Roccellin contained traces of tin; Orange I and Waterblue contained traces of tin and zinc. The last-named coloring matters were therefore not pre] tared in proper state of purity. The authorities in Vienna examined four and rejected two colors. {Ibid., 1S98, p. 107.) The Swiss authorities exercised control over colors, after they reached the market, with the result that the authorities in Basle examined ten colors and rejected one. (Ibid., 1897, p. 292.) These facts, together with the knowledge derived by even the most superficial ocular examination of the 294 specimens received in the summer of 1907, played a very great part in the formation of the con- clusion thai control of quality, in order to be even reasonably effec- tive, must be thoroughgoing, and that colors must be excluded from I lie market until they prove themselves to be clean, rather than permitted promiscuously and then driven out of the market by the authorities if unsuitable. The effective quality control of food colors requires careful and searching examination of a kind which can not usually be obtained LEGAL ENACTMENTS. 41 by the general purchasing public. The quality of the food colors offered in the summer of 1907 varied greatly, and the substances contaminating them were of such indefinite and probably variable composition (of whose physiological action nothing definite was known and whose quick and certain detection in the colored food product would be very complicated, if not impossible) that quality control of greater efficiency than that exercised by those selling food colors in the summer of 1907 seemed necessary on the part of the authorities having charge of the enforcement of the food and drugs act. The points of original entry of food colors into the United States food-color market are relatively few, whereas the points of distribu- tion of food colors are very many, the former being less than 20 and probably fewer than 10, while the latter may number up into the hundreds; therefore, not only is the labor and the expense of quality control of food colors reduced to its probable minimum by keeping food colors off the market until they have shown their right to be so used, but also the certainty and the efficiency of quality control is increased to its probable maximum. The quality control thus suggested is similar to that exercised by the States of New York, Michigan, and Ohio over salt before it enters the market for human consumption. The method of color control here suggested differs only in degree, not in kind, from the quality control exercised over salt by the States named. Experience has shown such quality control of food colors to be not only practi- cable but capable of realization without any hardship and but Little, if any, inconvenience to those concerned. STATE LAWS PROHIBITING THE USE OF COLORS IN CERTAIN FOODS, 1909. The laws of the individual States of the United States have also restricted the use of coal-tar coloring matters in foods. These restrictions are directed principally against the use of color to con- ceal inferiority, which restriction is found in almost all the States. The sale of poisonous coloring matters for foods is prohibited in the State of New York, and in New York and North Carolina the addition of injurious colors to foods is prohibited. Minnesota and North Carolina prohibit coal-tar dyes in all foods. Foods and beverages are considered adulterated in North Dakota and Wyoming if they contain aniline dyes or other coal-tar dyes. Artificial coloring is prohibited in sausages bj Colorado and Wisconsin. Artificial coloring, including, of course, coal-tar color-, must not he added to vinegar in the States of Arkansas, California. Connecticut. Iowa, Minnesota, Missouri, New Jersey, New fork, Pennsylvania, Tennessee, Wisconsin, and Wyoming. 42 COAL-TAR COLORS USED IN FOOD PRODUCTS. Distilled vinegar must not contain artificial color in Ohio and Oklahoma, and must be free from harmful artificial coloring matter in Utah. In South Dakota oleomargarine must not be colored. Artificial coloring is prohibited in milk by California, Oklahoma, Pennsylvania, Utah, and Wisconsin and in cream by California Connecticut, Pennsylvania, Utah, and Wisconsin. Coal-tar dyes are inhibited in cakes, crackers, candy, ice cream, and like products by Virginia. Ice cream is considered adulterated in Michigan if it contains harmful colors. Forty-six States prohibit the use of poisonous colors in candy. They are as follows: Alabama, Arkansas, California, Colorado, Con- necticut, Delaware, District of Columbia, Florida, Georgia, Idaho, Illinois, Indiana, Iowa, Kansas, Kentucky, Louisiana, Maine, Maryland, Michigan, Minnesota, Missouri, Montana, Nebraska, Nevada, New Hampshire, New Jersey, New York, North Carolina, North Dakota, Ohio, Oklahoma, Oregon, Pennsylvania, Philippine Islands, Porto Rico, Rhode Island, South Carolina, South Dakota, Tennessee, Texas, Utah, Vermont, Virginia, Washington, Wisconsin, and Wyoming. VI. RECOMMENDATIONS BY ASSOCIATIONS AND INDIVIDUALS AS TO USE OF COAL-TAR DYES AS FOOD COLORS. CAZENEUVE AND LEPINE. Cazeneuve and Lepine (Bull, de Vacad. de medicine, April 27, 1886, p. 643) says: We have arrived at the following conclusions: 1. The nitro derivatives are especially poisonous (dinitronaphthol being comparable with picric acid), but the sulphonated product is harmless. 2. Safranin and Methylene Blue are harmful, producing gastric intestinal dis- turbances, being violent poisons. 3. The following coloring matters are tolerated by man, whether well or affected with ISright's disease; similarly, too, animals (dogs, guinea pigs) without any noticeable disturbances and al rather high doses: Probable 1 trees Table Nos. Probable Green Table Nos. 5. Ponceau 11 55 6. Orange 1 85 7. Fuchsin S 462 1. Fast Yellow 9 2. Etoccellin 102 ;. Bordeaux B 65 1. Purple KM) or 107 Among the nontoxic Bulphonated colon we have been able t<> make out the following li-t <-r classification, baaed upon their power of producing disturbances, proceeding from tie- least Ineri to the meet inert: Probable Qreen i able Nos. ... Yellow NS 4 6. Fast Yellow 9 7. Purph' 106 or 107 Probable i i able Nos. 1. < hrange 1 86 2 1 '>' >rdeaux l '» . 68 A. Pom eau R 55 4. •llin 102 RECOMMENDATIONS BY ASSOCIATIONS, ETC. 43 The results of our experiments have led us to the following conclusions: 1 . The relative nonpoisonous nature of the azo colors used for coloring wines explains why this artificial coloring has not caused any real epidemic. 2. This artificial coloring of wines by coal-tar colors is dangerous. It opens the door to the employment of coloring matters of very variable and noxious properties. Thus Martius Yellow, which is poisonous, has been used for 10 years past to color pastry (3 grams per 100 kilograms) and it may be used to-morrow, perhaps, to color wines mixed with a red or a blue. 3. A rigid law against the artificial coloring of wines ought to be promulgated, particularly if this coloration covers detestable practices most prejudicial to the public- health. The addition of salicylic acid, glycerin, and tartaric acid, or the acidifying by sulphuric acid, is cloaked by the use of the coloring matter. Some sulphonated azo-coloring matters are sufficiently inert to enable their being employed as artificial color in foods, bonbons, and liquors. These colors are manu- factured according to simple processes which give theoretical yields and no metallic salt, such as mercury, tin, or arsenic participates; Btllpha is the only impurity. In view of the great extent of the use oi these coloring matters, it is better to regulate their consumption by tolerating certain of these products rather than to interp' illusory barrier to their use. 'Where you can not arrest a stream you can at least regu- late its coui It would be better definitely to classify these substance.- with respect to their noxious properties, tolerate some and prohibit the others, rather than to be exposed to the consequences of permitting manufacturers to introduce into food, without any scien- tific control whatever, any products whatever. These coloring matters should be sold in commerce under the names of harmless colors as determined by analysis. By chemical analysis it would be recognized as to whether we were dealing with one color or with a mixture of two or three colors. The colors most used are made up of red, yellow, and blue, which apparently imitate the appearance of the wines of the Midi. Thus we have recognized such coloring matters in Roccellin, Xaphthol Yellow, and Methylene Blue. Sulphonated Fuchsin is very much used, combined with a yellow and a blue. This mixture turns green with ammonia, like the coloring matter of wine. In fact, A. id Fuchsin is decolorized by ammonia. The yellow and blue remain intact, and give a green which suggests true wine color. SOCIETY OF SWISS ANALYTICAL CHEMISTS. In 1891 the Society of Swiss Analytical Chemists recommended that certain coloring matters which are to be regarded as harmful to health should not be permitted to be used in the preparation of articles of food intended for sale in which artificial coloring is at all permitted. The coal-tar colors thus prohibited arc identified in the following by their Green Table numbers, only one trade name being given: Picric Acid (l); Dinitrocresol (2)j Martius Yell-. Aurantia Orange II (86); Metanil Yellow (95); Safranin (584 ; Methylene Blue (050). (Zts. Nahr. Unters. Hygiene, 1891, v. TSCHIRCH. In L893 Tschirch recommended as follows: 1. Tie- 00tl-tU OOloiB, and in a narrow MOM the anilin ColoiB, arc DO longGf harmful on account oi then* arsenic content, since at the present time ■ majority of them arc prepared free from arsenic. 2. Some colors have shown themselves to be harmful to the system. 44 COAL-TAR COLORS USED IX FOOD PRODUCTS. 3. Coal-tar colors in general should therefore be permitted for the coloring of foods, but those that have been found to be harmful should be expressly and specifically forbidden. 4. The amount of coloring matter which has been determined quantitatively in bonbons and liqueurs is so small that even the ones regarded as poisonous would not be able to develop their harmful effects. (Zts. Nahr. Unters. Hygiene Waarenk., 1893, v. 7, p. 338.) KAYSER. In 1895 Kayser expressed liimself as follows : As to the poisonous nature of organic coloring, and in respect to their composition, H. Erdmann (Pharm. Centralh., 1892, v. 33, p. 357) concludes that in general acid iffs can pass as nonpoisonous; whereas in the case of basic coloring matters it is recommended to make a physiological examination before using them for the col- oring of things in daily use, especially articles of food. ^Tiether that portion of that view which deals with acid dyestuffs will retain unexceptionable and positive validity appears doubtful. At the present time, speaking generally, interested manufacturers take the point of view that all the coloring matters which are not forbidden in the food law are to be regarded as permitted. Whether this point of view is free from legal objection can not be discussed here; that, however, it can not be accepted from a hygienic point of view under any circumstances whatever does not require any special proof for those conversant with the facts. The hygienic requirements under all circumstances can be summed up in the following rule: Every coloring matter is to be regarded in every way as suspicious, so far as its harmlessness is not proven by experience or by correct physiological experiments. No one can say in advance that among the colors which are to-day manufactured and used, which are as yet not called into question, there are none which p distinctly poisonous properties. Correct examination of artificial coloring matters in this direction is, as is well known, even to the present almost wholly lacking. (Forsch- ungsberichte iiber Lebensmittel, etc., 1895, Vol. II, p. 181.) WEYL. In 1896 Weyl expressed himself as follows: Since the number of the organic coloring [natters already known is a very large one, and since their number is increasing daily, and it seems to be unlikely thai cadi individual of these coloring matters will be examined as to its poisonous nature, there an' only two ways left in which to solve the question :.- to the use of coloring matters in the manufacture of food and articles of daily use. One of them, and at the same time the simplest, would be to prohibit the use of all coloring matters for the coloring of foods, etc. This rigorous point of view will hardly be taken by legislators, because it would he tantamount to the removal of many marks of differentiation which have become desirable and necessary. The second way seems to he (he much more practicable, and which Theodor Weyl proposed some I ime i 1 1 con i i-^ iii permit ling the use of only a definite number of coloring matter.-, recog- nized ;i- harmless, for the coloring of articles o! food, etc. Which coloring matters • I.-- bo permitted is to he determined by the authorities having jurisdiction. 'lie- same authoritie are :>i o to determine the maximum amount of each coloring mailer uhi< h c.m he used for any purpose. New coloring matters can be used <>nly for the above-mentioned purposes when they have been recognized a- nonpoisonous after official test. All permitted coloring matters must he also detectable } even in small amounts. (Httndhuch r llijtjiate, 1896, Vol. Ill, p. $85.) RECOMMENDATIONS BY ASSOCIATIONS, ETC. 45 NATIONAL CONFECTIONERS' ASSOCIATION. In 1899 the National Confectioners' Association of the United wStates issued an official circular, which has been previously discussed, 1 designating certain colors as harmful, and certain others as harmless; the members of each class are given in the following list, in which only one trade name is given, the Green Table number appearing in parentheses at the end of that name. Harmful Organic Colors. Red colors: Ponceau 3RB (163); Crocein Scarlet 3B (160); Cochineal Red A (106); Crocein Scarlet 7B (169); Crocein Scarlet extra (164); Safranin (584). Yellow colors: Picric Acid (1); Martius Yellow (3); Acme Yellow (84); Victoria Yellow (2); Orange II (86); Metanil Yellow (95); Sudan I (11); Orange IV (88). Green colors: Xaphthol Green B (398). Blue colors: Methylene Blue BBG (650). Brown colors: Bismarck Brown (197); Vesuvin B (201); Fast Brown G (138); Chrys- oidin (17, 18). Harmless Organic Colors. Red colors: Artificial Alizarin and Purpurin (534); Eosin (512); Erythrosin Rose Bengale (520); Phloxin (521); Ponceau 2R (55); Bordeaux B (66); Ponceau 2G (15); Fuchsin S (462); Archil Substitute (28); Orange I (85); Congo R. Azoru- bin S (103); Fast Red D (107); Fast Red (105); Ponceau 4GB (13); Fuchsin Yellow and Orange colors: Xaphthol Yellow S (4); Brilliant Yellow (5); Fast Yellow (8); Fast Yellow R (9); Azarin S (70); Orange (43). Green colors: Malachite Green (427); Dinitroeoreeorcin (394). Blue colors: Indigo (689); Gentian Blue (457); Couplers Blue (600). Violet colors: Paris Violet (451 : Wool Black (166); Azoblue (287); Mauvein Brown colors: Chrysamin R (269). SCHACHERL. Schacherl in 1903 made the following statement : If coal-tar colors are to be permitted for the coloring of food, then, in my opinion. it is not right to limit the use of such to a few coloring matters, but groups of coloring matters must be permitted which are without suspicion from a sanitary standpoint, and which are characterized by definite reactions. Other groups, on the other hand. which contain harmful or merely suspicious colors, must be absolutely excluded. * * * The selection would be easily made if sufficient data were at hand with i to the physiological action. Unfortunately this is lacking, a circumstance which oeed do! be surprising in view of the very large number of synthetic coloring matters, since the Schultx-Julius tables enumerate 881 such colors. Unfortunately the experiments of Th. Weyl, which were planned on a I have not completed, and apart from isolated mvestigations we are limited in passing judgmenl upon th<' most of these colors to the proof that to date nothing with respect to harmful action has become known. * * * Consequently it is still possible that one or the other coloring matter which may to-day be regarded as ab ion, or i newly ..•red coloring matter may possesn poisonous properties; the 1- mould under all circumstances have the power to exclude for use in foods all suspicious combinations, and all such coloring matters as are not easily distinguishable from them. 46 COAL-TAB COLOBS USED IX FOOD PBODUCTS. The final recommendations of Schacherl amount to permitting — 1. All the Azo colors, Nos. 7 to 393 of the Green Tables, except No. 86. 2. All the Triphenylmethane colors, Nos. 427 to 492 of the Green Tables, except the hydroxyl derivatives, which would be Green Tables Nos. 433, 484, 485, 486. and 491. 3. All Pyronins, Nos. 493 to 527 of the Green Tables. 4. All Oxyketones, Nos. 537 to 570 of the Green Tables. 5. All Indulins, Nos. 599, 601, and 603 of the Green Tables. 6. Naphthol Yellow S, G. T. No. 4. 7. Methylene Blue, G. T. No. 650. The use of all other coal-tar colors would best be forbidden, partly from hygienic and partly from practical considerations (rendering control more easy), until the absolute harmlessness of the group in question is determined by physiological test. * * * It should be required of all permitted coloring matters that they shall not contain substances which are harmful to health, or even suspicious, either in chemical union or as contaminations. (Fifth International Congress of Applied Chemistry, Ber- lin, 1903, Vol. IV, pp. 1041-1048.) The exclusions recommended are all nitro-colors, except Naphthol Yellow S; all acridin colors and all chinolin colors; Auramin, In- dophenol and all nitro-colors, except Naphthol Yellow S, Schacherl regards as not necessary; further, he has no knowledge of the physio- logical action of any of the azoxy or the thiobenzenyl colors, and aside from Methylene Blue, he has no knowledge of the physiological action of the oxazins and thiazins. CLASSIFICATION OF RECOMMENDATIONS IN THE LITERATURE. The following table shows the groups of coal-tar colors of the Green Table classification and the Green Table numbers of the members of each of the groups, together with the favorable or unfa- vorable recommendations found in the literature in regard to each and a statement as to those regarding which no recommendations are made: Tabulation of recommendations found in the literature. Group num- ber. Green Table number. Name of color. Number of Individuals reported on— Total number. Un- favor- ably. (i) Favor- ably. (2) Con- flict- Injdy. (3) Re- port cl. (4) Not re- ported. 1 1-6.... 7 L32 . 390 U6.. 417 124.. 425 126.. 679 816.. 617 646 . 649 667 . ■ Nitro.. 4 3 3 ■> 10 8 is 7 6 83 13 2 M'iip>a/.o 94 :* l'.H 4 16 11 l l 1 II 2 1 8 7 Btllbene 17 H 8 1 r> •> 2 (i 2 2 2 a i i .■ .") 1 1 ! (t 8 i 1 •_) l 1 1 is 8 2 1 3 7 9 4 1 1 2 2 1 10 is 11 12 Xanthi 27 1 13 1 1 Anthracene 88 us 81 80 17 B 5 1 | 3 Sulphlda L9 21. 5 Total 33 32 41 106 689 RECOMMENDATIONS BY COLOR INDUSTRIES. 47 CONCLUSIONS. Applying the Schacherl rule, "Other groups which contain harmful or merely suspicious colors must be absolutely excluded," to this table and assuming that all entries in columns 1 and 3 shall be regarded as rendering such colors as "harmful or merely suspicious," it will be found that only one group, namely the Stilbene group, would be permitted under that rule; further, that tins rule would admit 17 colors, not one of which has been reported on in the litera- ture as to its physiological action. This state of affairs tends to em- phasize the difficulties in the way of any generalization which will be safe so far as public health is concerned and fair to those who use food colors for admittedly legitimate purposes and to make the following recommendation appears to be the only satisfactory way of solving the food-color problem : Although it would be possible to draw quite reliable conclusions as to the advisa- bility of employing certain colors for food products on the basis of their chemical constituency, the mode of their manufacture and of the ingredients used in same, nevertheless, I think that by far the safest way would be on the one side to force the dealers of colors intended for food products to sell only such colors with which exhaustive and careful physiological tests have been made by experienced and espe- cially impartial and thoroughly reliable people, thereby establishing their harmless- ness beyond a doubt. On the other hand, the manufacturers and canners of food products of any description should be forced to purchase and use only those colors which they are sure have been submitted to such careful tests as have been described and by these testa found to be harmless. (Licbcr, The use of coal-tar colors in food products, 1904, p. 150.) This view is confirmed by Santori ' (Moleschott's Uniersuchungen, 1895, Vol. XV, p. 57), who says: From all these experiments it follows that it is impossible, as some have desired to do, to conclude simply from the chemical composition and constitution whether any given coal-tar dye is poisonous or nonpoisonous. Tims [ndulin belongs to the same group as Printing Blue and Methyl Violet to the same group as does Acid Violet; therefore each individual coal-tar dye must be separately examined, and it is only by this laborious method that tie 1 use of all really poisonous coal-tar dyes will be prevented. VII. RECOMMENDATIONS MADE BY UNITED STATES COLOR INDUSTRIES AND TRADES TO THE DEPARTMENT OF AGRI- CULTURE. Prior to the issuance of any regulations, the commission on rules and regulal ions f'<>r the food and drugs act . June 30, 1906, held meet- ings in New York City during September of thai year. The steno- graphic reports of those meetings, and the briefs Bled, in so Far as ihev 'Santori examined 1.' dilTeretit Hue and violet dyes on dop by the month and hypodermic*] the month and 7 OOUS by the mouth, lie found Indulm to ' onous and Printing Bhn to be nonpoJeooooj; ktU Vlotol to be no np oiei poisonous. 48 COAL-TAR COLORS USED IX FOOD PRODUCTS. relate to colored food, colored food products, or material for coloring foods or food products, have been condensed verbatim in the following pages with the sole exception of omissions, as indicated ; for obvious reasons the names of those making the suggestions hereinafter quoted are not given. This review of opinions expressed and recommenda- tions given by the industries and trades most interested in the manu- facture, use, and sale of food colors and colored food products is believed to be fair and full with respect to each and every quotation. The numbers in parentheses following each quotation refer to the pages in the stenographic minutes of the hearing from which those quotations are made. ANTAGONISTIC TO ALL ADDED ARTIFICIAL COLOR. 1. Our position in the coloring question is that we are opposed to all artificial coloring matter in food products. * * * (p. 109). 2. Secondly, speaking first of our own business, and I believe that my view would hold as regards all food products, it is my opinion that all added artificial coloring matter in food products should be prohibited. My experience in our line of business demonstrates this to me beyond any question of doubt (p. 439). CONCERNING RESTRICTIONS AND REQUIREMENTS. WHAT CLASSES OF COLORS SHOULD BE PERMITTED TO BE USED. 1. All colors, irrespective of their class, whether animal, vegetable, or synthetic, which have been physiologically and chemically examined, and which will neither retard digestion nor have special physiological effects when consumed in quantities corresponding to two grains per day per adult (p. 106). 2. On the use of colors we recommend that any kind of a harmless color should be permitted provided it is not a color generally known to be poisonous, or generally found to be poisonous, or one that may be almost impossible to be produced without containing some poison within itself, when finished and ready for use (p. 119). 3. Only such colors as are guaranteed to be harmless by reliable manufacturers should be used in the manufacture of confectionery (p. 555). WHY COAL-TAB COLORS SHOULD NOT BE BARRED. 1. Coal-tar colors, as a class, should not be prohibited; but all those coal-tar colors generally found to be poisonous, or which are hard to produce without containing poisonous properties when ready ionise, should he forbidden the privilege of being used or offered for sale for use in food (pp. lit;, 117). 5. I ought to put in a plea for the use of coal-tar colors, harmless, of course, for the i that we have nol as yet been able to find any vegetable coloring that is suitable that will give us the results that we require. Goal-tar colon, as everybody knows, iich stronger and are more soluble and nol acted upon by acids, whereas the table colors, with but one exception, which is a dark red, we have found great difficulty in making use of fox bottled soda water. Almost all the vegetable colors either fade out or change on accounl of the citric acid in the syrup or food, or form in a little while B precipitation which renders the foods unsalable (pp. 1 L9, 120). 0. * * * aniline butter colorings * * * are superior to all vegetable colon in the following points: (1) Shade and brilliancy. (2) Strength, by which less foreign material is introduced into the butter. (3) Permanency when exposed to Light and cold storage. I i) No effect on the taste or flavor. (5) A clear solution RECOMMENDATIONS BY COLOR INDUSTRIES. 49 without sediment or mud which gives the butter a uniform tint without specks (pp. 176, 177). 7. We recommend * * * that the use of harmless coal-tar colors, such as chrysoidine, tropseoline, azoflavine, rocelline, ponceau, Bordeaux, Biebrich red, sulphonated fuchsin and naphthol yellow S be allowed, subject to declaration on the label of the quality and quantity of the color used (p. 226). RELATION OF LABEL TO COLORED FOOD PRODUCTS. * ■* * if it seems wise in the opinion of the commission to allow certain color- ing matters in food products, then the names of the coloring matter ought to be stated on the article (p. 109). 9. * * * I suggest that the use of aniline colors should be made proper if it is so stated on the label (p. 127). 10. We recommend * * * that the use of harmless coal-tar colors * * * be allowed subject to declaration on the label of the quality and quantity of the color used (p. 226). LABELING OF FOOD COLORS AS DISTINGUISHED FROM COLORS FOR OTHER USES. 11. We recommend that if by any means the Government has the power under the food laws to compel color manufacturers so to do, they be compelled to label all packages containing colors intended to be used in articles of food as colors intended for such purposes as distinguished from colors intended for other purposes; * * * (pp. 555, 555a). THE TEST OF A HARMLESS COLOR. 12. A harmless color is one "which will neither retard digestion nor have special physiological effects when consumed in quantities corresponding to two grains per day per adult" (p. 106). QUANTITY OF COLOR TO BE CONSIDERED IN DETERMINING HARMLESSNESS. 13. The quantity of 2 grains is mentioned here because in confectionery where these harml i - colon are more used perhaps than in any other product, it would be a proportion of one part of color to 3,500 parts, representing 1 pound of color to pounds of confectionery, and that is why that was accepted, because that is practi- cally the maximum quantity used in confectionery (p. 106). 14. In the very minute quantities in which the colors are used in carbonated bev- -. it would seem a great hardship to prevent us from using coal-tar colors. One ounce of coal-tar red will color satisfactorily from l.dOO to 2.000 gallon- of soda water. < >i yellow and orange 1 ounce will color from 1,000 to 8,000 gallons. It is readily leen that unless the coloring used is absolutely a violent poison it can have abso- lutely no effect on the consumer who takes it in an frounce glass, and who could not possibly consume ball • gallon or a gallon of that product (p. l L9 . 15. On this basis 1,000 pounds of butter would contain 420 grains of aniline color lc < foe ounce of (dor to 30 pounds colored food ip. 135). ■i hi MMiilli in on I I ki \ I COAL-TAB COD [BED. 17. And if I were on this committee I would advocate taking out two or three mi- which beyond any shadow of doubt are harmless, and which have been experimented upon, and which would !»«■ Nuflk-iem for all the purpose- nf the industry * * * none of these colon are patented, anybody can manufacture them (pp. 147, I ■ i Bull. L47— 12 4 50 COAL-TAR COLORS USED IN FOOD PRODUCTS. 18. You have been told that the food commission of the State of Pennsylvania is going to rule that seven colors shall be allowed in the State of Pennsylvania. That is not enough, because it is not possible to reproduce all of the required shades with those colors unless you take the seven primary colors, when you can reproduce any colors. But it is absolutely necessary to have about three yellows to meet the require- ments of the trade. Some of the articles that are put out have an organic acid nature, and the anilines will stand that. * * * On that account the character of the food product must be taken into account in the use of the color, and the confectioners have about 30 colors that I know positively of * * * (p. 160). 19. We recommend that * * * chrysoidine, tropaeoline, azoflavine, rocelline, ponceau, Bordeaux, Biebrich red, sulphonated fuchsin, naphthol yellow S be allowed * * * (p. 226). MANUFACTURING REQUIREMENTS WHICH COAL-TAR COLORS MUST FULFIL. 20. Coal-tar colors, as everybody knows, are much stronger and are more soluble and are not acted on by acids. * * * Almost all the vegetable colors either fade out or change on account of the citric acid in the syrup or food, or form in a little while a precipitation which renders the goods unsalable (p. 119). 21. There are many of these colors (coal-tar colors) that will not stand the natural acids produced in manufacturing foods. For instance, in making confectionery, in boiling candy you make a certain grade of candy where the mixture is boiled to 230°. A certain color will stand that temperature. Then you take another candy and that is boiled to 320° or 340° F., and the colors that will stand 230° will in many cases not stand the temperature of 340°, while another class of colors will stand that temperature. So you have got to distinguish and get a color that will stand these difficult requirements (pp. 159, 160). GUARANTEES AND GOVERNMENT CONTROL. 22. Only such colors as are guaranteed to be harmless by reliable manufacturers should be used in the manufacture of confectionery (p. 555). 23. We recommend * * * that the Government procure samples of such colors from time to time, wherever they have jurisdiction so to do, and if such colors be not legal for such purposes under the food law, that the 6ame be prosecuted and driven off the market in so far as the Government has the power to do so; and if no means can be devised to compel such labeling of colors intended for food proposes, then that the Government procure such evidence as possible as to the purpose for which a color is intended to be used, and if such evidence shows a food purpose, that such color be prosecuted if illegal under the food law (pp. 555, 555 a). 24. See page 13. 25. The chemical test is the first, and that, might throw out a color on account of its containing a little tin or zinc, or some substance foreign to the food product. Whether thai is deleterious in the quantity in which it is present or not is immaterial (p. 160). 26. In a brief hied the following suggestions were made: It must be stated that all the chemist can determine is whether or nol the colors contain some impurities that are known to be of poisonous nature. Aniline colors are or can be made entirely free of such impurities, and with this fact established the task with the chemist Is exhausted. When it comes to decide the question whether or not a color by itself, when free Of all impurities, is injurious to health or not, then the chemist is not the proper authority; it is for the physiologist and fox the medical pro- tops i on such questions. Chemical theories go for nothing in deciding such questions. It, would ool even do to classify colors or other substances according to their makeup, ai it has been shown again and again thai substances belonging to the same chemical class are entirely different, In regard to the physiological conduct. INVESTIGATIONS ON HAKMFULNESS. 51 Reliable information on this subject can be gained only by physiological experiment, as we can not say definitely whether a substance, color or any other, is injurious or not without finding out for every substance by experiment. This has been done for a considerable number of aniline colors, and these experiments are the only things that deserve any attention. Everything else is idle talk. * * * There are a very large number of aniline colors that have not been treated yet, and we are safe in saying that among these will be also some harmless and others injurious. As they have not been experimentally tested, we do not know which are harmless and which are not; It will therefore be clear that a law forbidding the indiscriminate use of aniline color for the purpose of coloring articles of food is necessary and useful. But if the meaning of the law is to prevent only the use of injurious colors, as it appears to be, then the way to proceed would be very definitely outlined. Besides physiological colors, all such colors should be forbidden that have been found to be injurious and such aniline colors as have not been tested sufficiently. There will then be left over a number of aniline colors which have been proven by experiment to be entirely harmless, even if taken in large doses. The experimenters were quite impartial. They had no preconceived ideas, but started simply to find out the true state of affairs. Their reports are therefore very reliable, and it will not do to overlook or to ignore them. The colors that they found harmless can be considered perfectly safe, so much the more as the doses conveniently taken with food would be much smaller than the doses that have proven to be harm- less. These harmless aniline colors carry all the shades wanted in the food industry. The law should provide that one of these colors (or mixtures of these) must be used when a food article is being colored, because these few aniline colors are the only colors that can be considered perfectly safe as far as our present knowledge goes. Nothing should be left to guesswork or experimenting, as is the case just now. The colors that are permitted should be enumerated by their scientific as well as by their commercial names, and only such colors should be listed as permissible for coloring food products as have been proven to lie harmless, even in large doses. Provision should be made to insure the purity of the colors sold for coloring food; the manner of packing such colors and the labeling of same should he laid down clearly, and all colors now listed, aniline as well as physiological colors, should be strict ly forbidden. If the problem is viewed without preoccupation and prejudice, the facts given above will speak for themselves. VIII. INVESTIGATIONS, OTHER THAN ON ANIMALS, BEARING ON THE HARMFULNESS OF COAL-TAR COLORS. PFEFFER. Pfeffer, writing on the Absorption of Anilin Colors by Living Cells, summarizes liis results as follows: The relatively little poisonous Methylene Blue does damage protoplasm in a solution of 0.001 percent. Methyl Violet: This coloring matter i- not only stored up in the juices of the cell, hut i- also able t<> color the living protoplasm, and care is necessary, on account <*" the poic J nature <>t" the Methyl Violet, n> prevent damage; these caution.- air I upon solutions of 0.0003 bo 0.0000] | trength. Methyl Violet less poisonous than Cyanin. Bismarck Brown about as poisonous a- Methylene Blue. ETuchsin a- poisonou ■ as Methj lene Blue. Bafranin a- poisonoui as Methylene Blue. Mnh j L Oral | only t<> a i light Tropeeolin <>< " in <><>, and Rosolic Acid are not p 52 COAL-TAE COLORS USED IN FOOD PRODUCTS. Methylene Green as poisonous as Methyl Violet. Nigrosin as poisonous as Methyl Violet. Eosin (Tetraiodo fluorescein) kills in 0.1 per cent solution, but lets live 24 hours in 0.01 per cent solution. (Untersuchungen aus dem Botanischen Institut zu Tuebingen, 1886-88, vol. 2, pp. 179-331.) WINOGRADOW. Winogradow reports on the influence of certain coal-tar colors on digestion, which experiments were carried out in glass. The con- clusions arrived at are as follows : ! The twelve colors, Safranin (584^, Azo Fuchsin G. (93), Coerulein S. (527), Jodeosin (516), Magdala Red (614), Benzopurpurin (277, 278, 279, 307), Ponceau 2R (55), Orange II (86), Phloxin RBN (?), Chrysanilin (532), Azoflavin (92), and Cerise (mixture of 448 and 532), even in amounts of a few milligrams, which in relation to the digestive fluid make up only a few tenths or hundredths of a per cent, exercise a strongly retarding, almost completely inhibitive, action upon the peptic digestion of albumen. The thirteen colors, Chinolin Yellow (667), Acid Green (434, 435), Azo Acid Yellow (92), Naphthol Yellow (4), Primulin (659), Anilin Orange (87), Metanil Yellow (95), Methylene Green (651), Iodin Green (459), Yellow T (84), Anilin Green (?) Auramin O (425), and Martius Yellow (3), retard the digestive action noticeably, although to a slighter degree than the first 12 colors; in any event they are not indif- ferent. (Zts. Ncihr. Genussm., 1903, v. 6, p. 589.) HEIDENHAIN. Heidenhain, in his book entitled "Ueber chemische Umsetzungen zwischen Eiweisskorpern und Anilinfarben" (Bonn, Germany, 1902), reports on the behavior of 70 different coal-tar colors, 3 intermediate products, and 4 raw materials toward various albuminoids such as serumalbumen, albumen, and casein. Of these 70 colors, 21 have been investigated physiologically, and the results embodied in this report; and of these 21, 17 were on the United States market as food colors and 4 were not on this market. Limiting the attention to the 17 that were offered, 4 of them are among the 7 permitted colors of Food Inspection Decision 76. In the cases where albumen and casein were used, they were employed in 0.5, 0.1, 0.02, 0.01, and 0.005 per cent solutions, acidified with acetic acid; the coloring matter was employed in a 1 per cent solution in the Hist strength, in 0.1 per cent solution in tliv second and third Strengths, and in a 0.02 pei- cent solution in the fourth and fifth strengths, and one volume of coloring-mat ter solution was brought in contact with five volumes of albuminoid solution. The following colors precipitated the albuminoid in all the strengths. The [lumbers in parentheses indicate the Green Table numbers; i ii i po Ible to Identify the trade n tmei given i>y Winogradow with specific numbers In kbles; the anxnben In parentheses after the Dame Indicate the number In the Green Tallies wherever that identification oonld be made with an; reasonable certainty; wherever two or more numbers ■ , the context Indicates that the name migbl apply to any one or all of them. INVESTIGATIONS ON HABMFULNESS. 53 the numbers in italics are those of the permitted list of Food Inspec- tion Decision 76: (55) Ponceau 2 R, (65) Fast Red B, (434) Light Green SF bluish, (56) Ponceau 3 R, and (106) New Ooccin. The following precipitated in all but the fifth strength: (107) Amaranth. The following precipitated onry the first three strengths: (14) Orange G and (85) Orange I. The following precipitated only the first two strengths: (462) Acid Magenta. For the basic colors the method of testing was different from that described for the acid colors and the amounts employed were not so definitely set forth. A 1 per cent solution of serum albuminoid was employed; the solution of coloring matter used varied in strength from 0.5 to 1 per cent (b), a "very dilute solution" (a) being also employed. The annotations given by the author (p. 114) are herewith reproduced in full : (17) Chrysoidin Y. (a) In a very dilute solution the yellow base is at once Liber- ated, and when sufficient color is added albumen precipitation takes place. No color change on heating, (b) Turns yellow at once. Further additional color prodi brown and albumen precipitation. (201) Manchester Brown, (a) Becomes yellow at once, due to separation of the free base. On heating no change, (b) Becomes a discolored brown, and prod nice brown albumen precipitate. (425) Auramin 0. (a) No change, (b) Precipitates albumen strongly. (427) Malachite Green. (The oxalate of the color was used.) (a) Color changes from a blue-green to a more pure green, (b) Cold, no precipitation of albumen; heated, sudden precipitation of albumen. (428) Brilliant Green. (A sulphate of the color was used Becomes milky. (6) Immediate precipitation of albumen. (448) Magenta. (Acetate and nitrate were used, and in both acted the -aim- Color changes from yellowish-red to rose-red. (6) Albumen precipitation abundant, ev percent a rum albumen; 5, I asein; the literature and the physiological action, as compiled herein, it will be observed that the ability to pre- cipitate albumen, or not to precipitate it, under the conditions <>!' Heidenhain, appears not to have anj direct connection with the results obtained by actual physiological test <»n animala or num. For instance, among the five thai precipitated all five Btrengths of 54 COAL-TAB COLOBS USED IN FOOD PBODUCTS. albumen and casein, No. 65, Fast Red B, of the Green Tables, has been found to be not harmful by tests actually described; No. 56, Ponceau 3 R, belongs to a class of colors genericalry permitted by the law of Austria; Nos. 55, Ponceau 2 R, and 106, New Coccine, have been reported on both favorably and unfavorably by different experi- menters; No. 434, Light Green SF bluish on physiological examina- tion has been described as suspicious. From this it appears that two colors, physiologically probably harmless, precipitated all the five strengths of albumen, and three colors, which are perhaps no more than suspicious, likewise precipi- tated all five strengths of albumen. No. 107, Amaranth, which has been examined with favorable results by two different experimenters, precipitates four out of the five strengths. Of the two colors precipitating the first three strengths, both hare been examined physiologically with favorable results, namely, Nos. 14, Orange G, and 85, Orange I. The color which precipitated only the first two strengths, namely, No. 4G2, Acid Magenta, has likewise been examined physiologically, and the reports are favorable. Thus it would seem that there is no definite connection between the physiological action and the ability to precipitate albumen from acetic acid solution in the case of acid colors. In the case of basic colors the situation seems to be somewhat different. OTHER AUTHORS. Rosenstiehl {Fifth International Congress of Applied Chemistry, Vol. Ill, j). 700) states that when the color is present in an excess, yeast absorbs 8 per cent of Magenta (448) and 5 per cent of Malachite Green (427, 428). The Acridins (528-538), the Thionins (649-657), the Safranins (583, 584), and the Rosanilins (447-448) dye yeast the best; solutions containing 3 per cent by weight of the dry weight of the yeast arc completely decolorized by such yeast and at ordinary temperatures. The Eosins and Phthaleins dye the yeast only incom- pletely, whereas Azo dyes (7 393) (with the exception of BenzO PuT- purin, 277, 278, 279, 309) do not dye yeast at all. Such dyed yeast, however, is not, dead; it has merely lost its power to cause fermenta- tion. The numbers in parentheses are the corresponding numbers in the Green Tables as nearly as they could he identified. Bokorny (CTiem. ZUj., 190(J,v.30,i).217) examined Magenta (448), Safranin (584), Victoria Blue (487, 488 or 490), Methylene Blue (660), and Alizarin Blue (562 or 563) (the numbers In parentheses are the probable Green Table numbers) as to their effect on micro-organisms such as yeast Cells, infusoria, and the like, and found that these dyes INVESTIGATIONS ON HARMFULNESS. 55 in a concentration of 1:100,000 killed them, whereas strychnin nitrate in the same concentration is substantially without effect. Death is caused by absorption of the dye by the albumen of the protoplasm. The dyes seem to be absorbed not only by living albu- men, but also by living nerve cells and fibers. Pure anilin or coal- tar colors, however, are not poisonous in the ordinary meaning of the words, that is, humans are not likely to be easily injured by them. Houghton (/. Amer. Cliem. Soc, 1907, v. 29, pp. 1351-1357) shows that Bismarck Brown (197, 201) and Crocein Scarlet (160?) hinder the peptic digestion of fibrin, casein, and albumen. Stilling (Anilinfarbstoffe als Antiseptica, 1890, v. 2, pp. 55-56) states that he found the animal cells to be affected by pure coal-tar colors in the same way that vegetable colls were affected. Penzoldt, based upon the experimental work of Beckh {Archiv. Exp. Path. Pharmak. 1890, v. 26, p. 310), reports as follows: Of the 15 dyes — Green Table Name. number. 1. Malachite Green 427 2. Hofman's Violet (methyl variety) 450 3. Methyl Violet . 451 4. Rose Bengal (Erythrosin) 517 5. Phenyl Blue 480 G. Methylene Blue 050 7. Fuchsin 448 8. Coralline 484 9. Eosin 512 10. Methyl Orange 8' 11. Vesuvin L97 12. Tropaeolin 88 13. Scarlet Red (?) 14. Congo Red 240 15. Indulin eulpho acid 001 all of which are water soluble and when used were free from arsenic, only the li in saturated solution arrested the development of Gtaphylococut pyog< r, and of these six all but the Erythroaii] and Methylene Blue arrested the growth of anthrax bacillus. Of these six when injected into rabbits subcutaneously the following produced no changes of consequence: Erythrosin (250), Phenyl Blue (100 . Methylene Blue The numbers in parentheses show the number of milligrams of dye per kilo body weight of the rabbit -. Methyl Violet (50) produced only local changes, sui h as gangrene. Malachite I Ireen LOO) and Bofman's Violet 1 20 : produced muscular pararj sis, which in the case of Malachite Green resulted fatally on the ninth day; in the case of Sof- man'a Violet the paralysis iras complete on the tenth day. The remaining nine colors are apparently «>f do effect upon staphylococui or upon anthrax II. \Y. Williams {A Manual of Bacteriology \ U \ p. M0), under " Disinfectants and Germicides," says: Aniline dyes. Many of these substances, notably pyoktanin (Methyl Violet . pos- ■ev germicidal properties. A. solution of 1:6000 will sill the anthrax bacillus in two 56 COAL-TAR COLORS USED IN FOOD PRODUCTS. hours. A much stronger solution, 1:150, is required to kill the typhoid bacillus in the same time. Malachite Green is said to possess even greater germicidal power than pyoktanin. Methylene Blue also possesses considerable germicidal power. IX. COMPILATION UNDER THE GREEN TABLE NUMBERS OF ALL INFORMATION AVAILABLE AS TO THE SUITABILITY OF COAL- TAR COLORS FOR FOOD. GENERAL STATEMENTS. Before entering upon a detailed study of what has been published for and against specific coal-tar colors, it is probably well to consider, for whatever they may be worth, some of the general statements that have been made, from time to time, in the literature relative to coal- tar colors, considered either as a whole or as subdivisions or classes thereof, and their physiological action or their suitability for use in foods. 1. Schultz (Die Chemie des SteinkoMenthcers, Brunswick, 1887-1890, Vol. II, p. 85), after discussing the regulations of the German Empire in respect to food coloring, says: With respect to these regulations the artificial organic coloring matters can be regarded, in general, as harmless. For the purposes of dyeing magenta made by means of arsenic acid, further picric acid, and those coloring matters which occur as oxalates or zinc chlorid double salts, such as Methylene Blue and Bitter Almond Oil Green, can be used. The use of the substances named for the coloring of food products is, however, suspicious and should not be permitted. 2. Stilling (Anilinfarbstoffe ah Antiseptica, Strassburg, 1890, Vol. II, pp. 55-56) says: In view of the fact that the most innocent substance, such as distilled water or com- mon salt, when introduced in large quantities into the organism can act fatally, the anilin coloring matters therefore, particularly if they be free from all admixtures, such as arsenic, copper, and chlorid of zinc, are to be regarded as wholly nonpoisonous. All experience gathered since my first publication has likewise fully confirmed this nonpoisonous nature. 3. Stilling (Ber. Klin. Wochensch. 1890, p. 531) also says: Proceeding from purely theoretical views, and based upon these botanical and physiological experiments, 1 have recommended anilin colors as antiseptics for the following reason-: J. They are to be designated as absolutely nonpoisonous. This will be confirmed by every chemist acquainted with the relevant details, and also from the medical side this has long ago been determined by Grandhomme. The publications of tins author, who made extern i ire observations and experiments in the anilin factory ot' Heister, Lucius & Bruening, appeared in the beginning of the ei| hi ies, and has hardly become known in medical circles. However, I was first made acquainted with this by my colleague, Prof. Rose, in Btrasburg. 2. That ii Is possible to bring about death in experimental animals by introducing large amount! of the coloring matter into the peritoneal i a\ ii >• proves nothing against COMPILED DATA UNDER GREEN TABLE NUMBERS. 57 the nonpoisonous nature of these substances. This action is to be regarded as a purely mechanical one, a view which I will thoroughly confirm in my second communication. With respect to the anilin colors not soluble in water, Ehrlich long before me, in his excellent publication on the oxygen requirements of the organism, has arrived at this view, and has excellently described it as merely a penetration of the organs. 4. Lelimann (Meihoden der Praktischen Hygiene, Wiesbaden, 1890, p. 54-3) says: The hygienic significance of coal-tar colors has heretofore been judged quite vari- ously. When the intensely poisonous nature of the first impure and particularly arsenic-containing coloring matters became known the inclination was to judge the coal-tar coloring matters very strictly; when it was subsequently recognized that the contaminations were largely the cau.se of the harmfulness to health, there followed a period in which no poisonous coal-tar coloring matter was known in any pure con- dition. (Eulenberg & Vohl, 1870.) More recent investigations, however, have disclosed a series of coal-tar coloring matters which, as a matter of fact, possess a con- siderable poisonous action, and already cases, although not numerous, have become known in which serious and even fatal poisonings by means of pure coal-tar colors have arisen. Alongside of this there still continue to exist the possibilities described by Eulenberg and Vohl (Viertel-Jahressch. fur Gerichtliche Mediz, 1870), whereby harmless coloring matters become harmful; but the realization of these possi- bilities has become essentially more seldom through improvements and chanj manufacture. 5. Stilling (Arch. Exper. Pathol. Pharmal., 1891, v. 28, p. 352), in speaking of the anilin colors as antiseptics, says: It is the nonpoisonous nature of these substances, their easy solubility and dif- fusibility, and above all their inability to coagulate albumen which lends them their importance, which now can be only difficultly denied. Note. — The work of Heidenhain abstracted in Section VIII does not fully bear out this article. 6. Erdmann (Pharm. d nfml/i., IS!' J, t>. 38, p. ps: The sulphonated as well as the carboxylated coal-tar dyes will not have any pro- nounced action on the organism. Acid dyea may therefore be regarded in : as nonpoisonous, whereas in the case of basic dyes a physiological examination is to be recommended before they are permitted to be applied to articles in daily use or indeed to be used in food or drink. Note.— Out of the 80 different dyes on the food-color market in the summer of 1907 whose composition was avowed, i;> were basic and 65 were acid. 7. Tschirch expressed himself as follows: 1. T and in a narrower sense the anilin colon, arc no I harmful on ao ounl of their arsenic content, since at the j ties* ml time the great majority of them are ; . free from arsenic. 2. Some colors ha\ e shown themselves to be harmful to the system 1 oal-tar col aeral, should therefore be permitted for the coloring of but those that have been found to be harmful should be expressly and specifically forbidden. 58 COAL-TAB COLORS USED IN FOOD PRODUCTS. 4. The amount of coloring matter which has been determined quantitatively in bonbons and liqueurs is so small that even the ones regarded as poisonous would not be able to develop their harmful effects. {Zts. Nahrs. Unters. Hygiene Waarenk., 1893, v. 7, p. 338.) 8. Georgievics (Lehrbuch der Farbenchemie , Leipzig, 1895 p. 10) says under "Poisonous nature of the coloring matters:" It is a little known fact that of the very large number of organic coloring matters only a few have been found to be poisonous; these are Picric Acid, Victoria Orange (Saffron-surrogate), Aurantia, Metanil Yellow, Orange II, and Safranin. The preju- dice which is still quite widely accepted that most of the artificial coloring matters are poisonous dates from the early periods of anilin-color manufacture, at which time magenta and the coloring matters made from it occurred in commerce highly con- taminated with arsenic. At present, however, these coloring matters are prepared absolutely free from arsenic, and are, as such, nonpoisonous. A few coloring mat- ters which occur commercially as zinc chlorid double salts, such as Methylene Blue and Malachite Green, may be harmful in consequence of their zinc content, and should therefore never be employed in the coloring of food products. * * * In consequence of their physiological activity some coloring matters are employed as medicines, indeed principally Methyl Violet, Auramin and Methylene Blue. The first two, known as blue and yellow Pyoctanin (pus-destroying) are, owing to their great antiseptic action and diffusibility, valuable medicines; on account of the unpleasant coloring effect accompanying them they are but little used. Methyl Violet was first recommended as an antibacterial remedy in diseases of the eye by Stilling; subsequently it has been employed in other special cases; its principal use, however, is in surgery for the prevention of malignant proud flesh. The use of Auramin is entirely analogous. Methylene Blue (as a free base) is used principally as an analgesic (pain-relieving remedy) and is given internally; on account of its ability rapidly to diffuse through the tissues of the nervous system; it can also be introduced by injection. It is used as a remedy against malaria, carcinoma, Bright's disease, etc. The following have been tested as remedies or as antiseptics: Safranin, Lydin (Mauvein), Vesuvin, Anilin Blue, Carbolic Magenta, Alizarin Yellow C (Gallaceto- phenone) , etc. The potassium salt of Dinitro-ortho-cresol was brought into com- merce by the Farbenfabriken vorm. Friedr. Bayer & Co., under the name of Anti- nonnin, and has given excellent results as a means against formation of mold in cellars and against wood fungi. 9. Weyl (Handbuch der Hygiene, 1896, p. 378) says: A few organic coloring matters, but only a very few, possess poisonous properties. A rule by means of which the poisonous or nonpoisonous nature of organic coloring materials can be determined without experiment is unknown even for those coloring matters whose constitution has been determined and experiments on the poisonous nut arc of organic coloring mat tor- an- very few in number. 10. Lewin {Lehrbuch der Toxicologie, 1897, p. 230) says: In the use ol various fabrics or of foodstuffs, which are colored with anilin or coal-tar colors, or in commercial contact with such colors, local and general symptoms ..!' poisoning, SUCh BJ eczema, swelling of the face, vomiting, diarrhea, amesthesia, paresis, etc., have been observed. The e are generally duo to the toxic nature of the coloring matters, frequently to harmful ingredients of the same, and hardly ever to poisonous mordants. Many workmen in anilin factories -how permanent spots; nmple, On the cornea ami conjunctiva, head, Chest) lace, ami neck, without COMPILED DATA UNDER GREEN TABLE NUMBERS. 59 any interference with their general condition. Local changes of more serious nature have more frequently been observed in the mucous membranes and on the skin. Thus, in one case, a camel's-hair pencil, soaked with anilin color, accidently entered the eye, and at first nothing was noticed but a violet blue coloration, later inflamma- ion and chemosis took place. I have observed local swelling and indurations of the skin, particularly on the cheeks, in the case of children after they had worn caps colored with anilin dye. 11. Winton (Connecticut Agric. Eiper. Sta. Report, 1901, p. 181) says : Although there is evidence that most of the coal-tar dyes are not injurious to some of the lower animals, it is not safe to assume that they are entirely harmless to human beings. The dog, an animal used in most of Weyl's experiments, has a proverbially strong stomach, and eats with no apparent discomfort many things which would disturb the digestion of a man. 12. Chlopin, in his book published in 1903 (see p. 75), state- i - follows : {Page 114.) * * * All the dyes examined by me I divided into three categories: Dyes which caused striking general symptoms of poisoning and led to the death of the animal, or would have led to it if the experiments were not purposely discon- tinued, I designate by the term poisonous; dyes which induced some separate and temporary symptoms of disease, for instance, vomiting, diarrhea, separation of albu- men in the urine, the general condition remaining normal, I designate as suspicious: lastly, the dyes which caused no apparent disturbance during the experiments are designated by the term nonpoisonous. I purposely do not call the last category harmless, because by our experiments the question could not be decided negatively as to whether the nonpoisonous dyes did not cause some finer pathological changes in the organism and functions which could not be detected by simple observation. (Pages 219-221.) Thus, according to all the investigators quoted, there were found altogether 22 poisonous and harmful dyes, out of about 60 dyes examined; which makes 30.7 per cent of poisonous and harmful dyes among those examined. My investigations gave 30 per cent of poisonous and 40 per cent oi suspicious The percentages above given have a fairly well established basis, since they were obtained by the examination of 100 dyes, which is about one-fifth ol all the dj commerce. Further, considering the distribution ol the poisonous and harmful dyes according to various chemical groups, we find that they occur in 12 oi the L8 groups, and we can not note any regularity in this distribution; ii is impossible to say that there is any definite connection between the feci that the dye belongs to a certain chemical group and its action on the animal organism. Usually among the .'. one ami the same group there are some harmful one-, but there are also some harmless ones, and the ones and the others have very similar composition. This or that action of the dyes on the animal organism, as we shall presently see, u d< termined m< the delicate < Inference in the internal structure of their molecules than by those differences <>n which i- based, at the pi of time, the classification of the aromatic dyes. On the I' .i-i~ of tin- whole experimental material on hand mine and that of other invest igatora I can make only the following very few and purely empirical generalisa- l . According to the shade produced, the poisonoui and harmful colors are distributed ai follows: Blosl "\ all poisonous dyes are found among the Yellows and the < » ranges; then come the Blues, then the I'-'", -ami the BUm k ry few harmful among the Violets and Qi found ool] picious one, and DO poisODOUi ones. 60 COAL-TAR COLORS USED IX FOOD PRODUCTS. 2. The most poisonous dyes belong to the Nitro, Azo, Tri phenyl, and Thiazin groups, and also to the Auramins. 3. A whole group of poisonous and harmful dyes is formed by the new sulphid dyes known as Yidal's dyes. (Page 224-) By the facts and observations quoted above is corroborated the opinion that in general the coal-tar dyes, according to the composition and properties, appear as substances foreign to the animal organism, and may influence harmfully the vital functions, even in those cases when they do not possess distinctly poisonous properties. For this reason many hygienists make it a principle not to allow the coloring of food products or of beverages with coal-tar dyes, independently of the fact whether they prove in actual experiments on animals poisonous or not. We must therefore agree with M. Rudner that food of the masses require the most far-reaching protection, maintaining them free from foreign additions. 13. Koenig {Die Menschlichen NaJirungs- und Genussmittel, Berlin, 1904, Vol II, p. 162) says: Even though the majority of the anilin coloring matters, in view of the small amounts in which they are generally employed, can not be regarded as directly harm- ful to health, yet the objections to their use in the coloring of food products for the purpose of substituting or strengthening a natural color lies in the deception con- nected therewith * * *. 14. Fraenkel (Arzneimittel-Synihese , Berlin, 1906, j). 570) says: It is clear that the coloring property of these chemical substances stands in no relation to their physiological actions, but, on the contrary, the physiological actions depend only upon the general structure of these substances, and therefore upon their membership in definite chemical groups. (Pages 574-5.) We see, even in considering this group of substances, that they do not possess any specific action, but they are capable of use, preferably by external application, as antiseptic materials, as materials which in their action Btand somewhere between carbolic acid and corrosive sublimate, and whose coloring property, in conse- quence of which they were primarily selected, is directly a hindrance in this use, since the coloring of the bandages and the hands of the operators and the skin of the patients certainly can not be regarded as a pleasant occurrence; that the antra ptic action stands in some relationship to the properties of the substances as coloring matter must be positively contradicted. It depends <>nly on the general structure of the substance, and does not stand in any direct relation to the ehromophore and anxo- chrome groups of the substances, but more closely to the aromatic nucleus. Indeed, it may happen that an auxochrome group diminishes the antiseptic activity of such a substance. Note. —The dyes referred to belong to the Monazo, Disazo, Triphenyl-methane, Xanthin, Azin, and Thiazin clas /'"'/< '>'..) The investigations of Ehrlich have shown that basic dyes color the brain gray, and, moreover, they color nerve fiber very well, and are therefore to be regarded as neurotropy . Thedyeacids, on the other hand, do not dye nerve fiber, and in particu- lar the at) itituted sulphonic acids do not dye tissue at all. ir>. Meyer, in his paper on "A preliminary communication on the toxicity of some aniline, dyes tuffs" (./. Amer, ( 'In iii. Soc.f t907, v. 29, p. 892 (Page 898.) "A manufacturing confectioner of this city, for whom I made examina- tion of colors used by him, informs mo that a yellow color sold as Anraniin has such COMPILED DATA UNDER GREEN TABLE NUMBERS. 61 high tinctorial power that 1 ounce will color 2,000 pounds of candy to the highest yel- low tint required in his business. It is obvious that the toxicity of such a body would have to be very high to render it harmful in such use." Conclusions of this kind do not take into account the possible detrimental action ensuing on healthy as well as diseased persons from long-continued use of small quantities of foreign substances. {Page 909.) The same author raises objection to feeding experiments on the ground that substances are thereby introduced greatly in excess of the amounts generally found in foods and that the ill effects "are liable to be due to the excess and in long- continued experiments due to a cumulative action of the excess." Surely if exc< amounts have a cumulative action, small amounts may also finally show toxic effects due to retention and accumulation of the poison. To declare a substance entirely innocuous would require evidence as to its nontoxicity both to normal and diseased persons after its long-continued administration in both small and large doses. The most extreme contingencies would have to be provided for. The above objections to feeding experiments are therefore not valid. It is hoped that a study of the effects on metabolism of some of these substances will help to further elucidate the subject. He summarizes the results of his physiological investigations of seven different coal-tar colors as follows: 1. Several commercial organic dyestuffs (Curcumin S, Tartrazin, Xaphthol Red 8, Carmois in B, Naphthol Yellow S, Gold Orange, and Ponceau 2 R) were studied ae to their general effects on dogs when administered in varying amounts and during fairly long periods (two weeks). 2. None of these dyestuffs under the conditions above indicated exhibited any marked degree of toxicity. There was only one fatal result, which may have been due to influence independent of the action of the colorant. Similar quotations from the literature could be added to the fore- going! but these are beyond question sufficient to show that a wide divergence of opinion as to the harmless or harmful nature of the coal-tar colors as a class does exist among scientific men, and that all those above quoted agree that there are some at leasl of the coal-tar colors which even in a pure state may be harmful to human health, and that the question of actual harmfulness under the conditions of actual use in foods and the consumption of foods Is regarded by some as being properly ans'svered in the negative and by others as being properly answered in the affirmative. The question of amount of color employed in t he food products and t he amounts of such food normally eaten are therefore raised by some as the deciding factors. In this connection the following statement from page 19 of the arguments before the Committee on Patents in the Bouse of Repre- sentatives, April 8, 15, 16,22,29, L908, may be of interest: * * * It should lx- remembered thai after a new chemical ha- been disc* and patented it requirei a- many u three year- oi experiment before we dare offer it in this country ai a medicine for human beings. These experiments are cond ah cud before we receive il here, li is first tried "n animal- and gradually, with caution, extended t>> human beings in tin- foreign hospitals, - certain its physiological effects quantitatively upon the various organs, both when i> an- in the healthy state and when they are affected by various disorders * * *. 62 COAL-TAR COLORS USED IX FOOD PRODUCTS. A search of the literature herein compiled fails to disclose any such searching physiological examination of any of the coal-tar colors recommended for use by human beings in food products, as is asserted in the above quotation to be necessary in the case of a new chemical intended for use as a drug. If Fraenkel, as quoted on page 60, is correct in his statement that coal-tar colors act physiologically because they are chemicals and not because they are coloring matters, then coal-tar coloring matters prior to use in foods, in winch they are used by the young and the old, the well and the sick, without restric- tion and without supervision, should also be thoroughly tested, and very few, if any, coal-tar colors seem to have been examined with the thoroughness set forth in the above quotation. That uniformity and purity of product is necessary in order to be sure that the chem- ical is going to act physiologically in the same way every time is obvious. According to Fraenkel, what is true of a chemical is just as true of a coal-tar color, and if uniformity of strength and cleanli- ness of product are desirable when a chemical is to be used as a medi- cine, such properties are equally desirable when a chemical is to be used as an ingredient in food. CLASSIFICATION OF OPINIONS IN LITERATURE AND IN LEGAL ENACTMENTS SHOWING CONDITION OF THE MARKET IN 1907. The literature and legal enactments hereinafter grouped under the relevant Green Table numbers have been classified as (1) unfavora- ble — i. e., only unfavorable reports found in the literature; (2) favor- able, and (3) contradictory reports, as each case required; so that under each Green Table entry there is not only the relevant literature, but also the character assigned to it for the purpose of coming to a conclusion as to the propriety of the use of such color in foods as based on such literature, which conclusion formed in that respect the basis for Food Inspection Decisions Nos. 76, 77, and 106. However, ii does not follow that all dyes placed in the "favorable'' list are actually harmless; the investigations or opinions reported of each may very well be based upon insufficient data. This classifica- tion, therefore, is merely intended to reflect the present state of the literature with respect thereto, and is not necessarily final nor con- clusii e. In substantially all the recorded cases the observers directly or indirectly asserl the absence of arsenic and mineral poisons in the dyes subjected to physiological best, but the kind of other impurity, if any, is not stated. For the purpose of a comprehensive survey of the literature and Legal enactments, the following tabulation is presented: COMPILED DATA UXDER GREEN TABLE NUMBERS. 63 Condition of the United States color market in the summer of 1907. [x=notoD market. Figures indicate number of dealers offering sample. Cross lines separate the several groups.] Green Table number. Unfa- vor- able. Favor- able. Con- tradic- tory. Green Table number. Unfa- vor- able. Favor- able. Con- tradic- tory. Green Table number. Unfa- vor- able. Favor- able. Con- tradic- tory. 1 X X X X 164 166 169 188 197 201 240 269 277 287 X 521 527 X X 2 X 1 i " 4 3 M 10 X 530 532 X X 5 6 2 1 ...... 563 X 8 9 5 1 2 6 2 X X 2 1 X X 2 X X 572 574 576 X X 11 1 13 X 14 394 398 X ...... 15 584 593 599 600 601 614 1 16 X X X 17 399 X 18 28 X 425 3 1 41 X X 43 427 428 433 434 435 448 450 451 457 459 462 467 477 478 479 480 483 488,490 2 3 55 65 2 l" 1 620 639 X X 70 78 X 4 ...... X 5 X 84 2 X X 649 650 651 .',.-,4 X 85 2 86 X X 87 88 X X X 89 "~6" 1 X X 2 X X 659 X 92 93 887 1 94 95 2 670 675 X X 97 1 X 102 X 6 1 "5" 103 X 689 105 106 107 3 502 504 512 516 517 520 2 7 5 Totals 106 *50 33 82 1 16 ...... 3 41 138 X X X 160 5 2 163 1 Italicized fijniros are colors in the permitted list, Food Inspection Decision 7f>. 2 On United States market In 1907. This table shows thai of the 106 coal-tar dyes examined physiolog- ically only 50 were on the United States market; Further, out of 33 " unfavorable" dyes S, or one-fourth, were on the United States market; out of 32 "favorable" dyes L6, or one-half, were on the United States market, and finally thai oul of 41 "contradictory" dyes 20, or very nearly two-thirds, were on the United States market in t lie summer of 1 907. Aflfliirmng this classification to be substantially and essentially fair, the only Green Table numbers which are of interest for the present discussion are those classified under "favorable/ 1 because any color positively injurious or of doubtful character is considered as being properly excluded from use in food products. The Green numbers classified as •favorable" are 32 in number, as follow 5; 28; 13; 102; 103; 106; 107; L< jfS6; W2; 467; -177; 512; 617; 520; 521 ; 527; 576; 59 64 COAL-TAE COLORS USED IN FOOD PRODUCTS. Of these 32, 16 were on the United States market in the summer of 1907 and their composition disclosed; they are presented in the following table, together with the number of dealers, out of a possible 12, offering them for sale. Distribution of "favorable" colors on the American market in 1907. Gr^en Table number. Number of dealers handling. Green Table number. Number of dealer? handling. Green Table number. Number 1 Green of dealers'! Table handling, number. Number of dealers handling. 4 65 85 89 10 2 2 103 105 107 169 6 1 7 1 240 433 435 462 1 1 4 ■2 512 517 520 692 3 5 2 3 From among these 16, six of the seven permitted colors of Food Inspection Decision No. 76 were selected. The process of selection and of elimination is described on page 166. CLASSIFICATION ACCORDING TO CHEMICAL COMPOSITION AND SUITABILITY. In the following table the chemical composition of the substances corresponding to the Green Table numbers is given, and their classi- fication according as the literature is regarded as being (1) unfavor- able, (2) favorable, or (3) contradictory in regard to the color. The colors are also arranged in the groups to winch they belong chemically, so that this table shows: (1) The number of groups reported on in the literature, (2) the number and composition of members of each group so reported on, and (3) the interpretation here placed upon such reports in literature. This table is given in the expectation that it will be of use to chemists and physiologists. Opinions as to suitability, classified according to groups and chemical composition. (The chemical nomenclature is that of the Green Tables; "a" is used for alpha and •• 1>" for bete.) T'nfa'. Favorable. Contradictory. NITRO COLORS. 1. Symmetrical frinitrophenol. 2, Dinitro o and p-eresol. .',. Diiiilro-a-naphlhol. fc Bexanltro-diphenylamliL Kfethyl-benzenyl-amldo-thlo- '.l-a/o-a-naphthol h onic add. 22. Wphenylamln yellow with iLitro-'liphriivlaiiuii. s. Aiiihlo-a/D-binzi'iic-'li- and monoeulpbonlo add, 2. AmldO - MO - toluene - disul- pbonlo add. 1 1. Bentcne aaO'b-paphthoi 18. Hen/i'iira/o-li- naphthol - b - BulpuoDk add. 1 1. Benxene4tfo>b-aaphthol dl- sulphonlc acid G. COMPILED DATA UNDER GREEN TABLE NUMBERS. 65 Opinions as to suitability, classified according to groups and chemical composition — Con. Unfavorable. mono azo colors— continued. DISAZO COLORS. 164. Sulphobenzene-azo-sul p h o - benzene-azo-b-naphthol - sulphonic acid. 201. Hydrochlorid of toluene dis- azo-m-tolylene-aiamin. 277. Ditolyl-disazo-binaphthion i c N1TROSO COLORS. STTLBENE COLORS. DIPHENYLMETHAXE COLORS. 4.'.',. J I ydrochlorld of lmido-tetra- methyl-diamido-diphenyl- methane. TRIPIIENYLMETIIANE COLORS. 434. Dimethyl -dibenzyl-diarnido- trlphenyl-carbinol-tris u 1- phoni 4.V.t. Chloric! of heptamethyl-rosan- ilin chlond. 478. Triphenyl-pararosanilin di- uri'l trisulphonic acid, tphenyl-pararosanilin-l r i- sulphonic acid. 488.490. Hydrochlorid of phe- n> 1-t • Diethyl- triamido-diphenyki-aaph- thyl carbinol. Favorable. 93. p-sulphobenzene-azo - dioxy- naphthalene sulp h o n i c acid. 102. p-sulpho-raphthalene - azo - b-anphthol. 103. p-sulphonaphthalene-azo-a- naphthol-p-sulphonic acid. 105. p-sulpho-naphthalene - azo - b-napbthol monosul- phonic acid. 107. p-sulpho-naphthalene - azo - b-naphthol-dis u 1 p h o n ic acid. 166. Sulphobenzene - azo-sulpho- benzene - azo - p - tolyl - b- naphthylamin. 169. Sulphotoluene-azo-foluene- azo - b - naphthol - a - sul- phonic acid. 240. Diphenyl - disazo - binaph - thionic acid. Contradictory. 394. Dlnitroso-resorclnol. 399. Azoxy-stilbene-di-sulphonic acid. 433. Diethyl- dlbenzyl-diamido- trlphenyl carbinoi-disui- phonic ;k - 1 < 1 . ).;.",. Iu.-thvl-ditH-nzyl-diamido- triplionyl-carbinol-tris u 1- phonicacid. sulphonic acid of lin and pararosanilin. ■177. Trlphfinyf- rowan H In aulphonJo acid and trlphe- ii> 1 - pararosanilin -mono- iulpnonio acid. 4».7. Dlsulphonicacld of dimethyl dibensyldi-ethy] trianudo triphehylcarbiuol. 15. Benzene-azo-b-naphthol di- sulphonic acid R. 16. Dimethyl-amido - azo - ben- zene. 17. Hydrochlorid of diamido-azo- "benzene. 18. Hydrochlorid of benzene-azo- m-tolvlene-diamin. 41. Hydrochlorid of toluene-azo- m-tolylene-diamin. 43. Toluene-azo-b-naphthol- sul- phonic acid. 55. Xylene-azo-b-naphthol-disul- phonic acid. 70. Dichloro-phenol-azo-b - naph- thol. 84. p-sulphobenzene-azo - resorci- nol. 86. p-sulphobenzene-azo- b-naph- thol. 87. p-sulphobenzene-azo-dimeth- ylanihn. 88. p-sulphobenzene-azo - diphe- nylanun. 95. m-sulphobenzene - azo-diphe- nylamin. 106. p-sulphonaphthalene - azo-b- naphthol-disulphonic acid. 138. Blsulphobenzene - disazo - a- naphthol. 160. Sulphobenzene-azo-benzene - azo - b - naphthol - niono- sulphonic acid. 163. Sulphobenzene- azo -sulpho- benzene-azo-b-naphthol. 188. Disulpho-b-naphthalene-azo- a-naphthalene - azo - b - naphthol-disulphonic acid. 197. Hydrochlorid of benzene-dis- azo-phenylene-diamin. 269. Ditolyl - disazo - bi - salicylic acid. 287. Ditolyl-disazo-bi-a- naphthol- p-siilphonic acid. : 108. N It roso-b-napht hol-b - mono - sulphonic acid. 427. Chlorid of tetra-methyl di-p- Idc-triphenyl-carbinoi. 428. Sulphate or chlorid of tetra- eihyl-diainido- trip!. LnoL 44s. Hydrochlorid i i Janilln and in. 4.v. llydrochlorids or a inoiio-di or trlmethyl (or ethyl) rosaolUna and para- '. ims. 451. Hydrochlorid Of ollin. 4.".7. Hydrochlorid. snip' triphenyl i lm and triphenyl pararoa- anilin. 1-rosantlin and (ri- . nilin sul- ,- rin kCid. 97291°— Bull. 147- 66 COAL-TAR COLORS USED IN FOOD PRODUCTS. Opinions as to suitability, classified according to groups and chemical composition — Con. Unfavorable. Favorable. Contradictory. XANTHENE COLORS. 502. Triethylrhodamin. 516. Diiodo fluorescein. ACRIDIN COLORS. 530. Ilydrochloridofdiamido-phe- " nvl-dimethyl-acridin. 532. Nitrate of chrysanilin and homologues. ANTHRACENE COLORS. INDOPHENOL COLORS. 572. Tin compound of dimethyl p-amido phenyl-p-oxy- a-naphthylamin. 571. Ilydrochlorids of p-pheny- lene-diamin-p-amidophe - nol and diamido-dipheny- lamin. AZIN COLORS. 602. Spirit Nigrosines. 014. Ainido- naphthyl- naphtha - zonium chlorid and dia- mido naphthyl-naphtha- zonium chlorid. ox AZIN COLORS. 020. Dimethylamido - dioxy - phe- nazoxonium carboxylate. G39. Chlorid of dimethyl-amido- naphtho - phenoxazonium chlorid. TIUAZIN COLORS. 640. Zinc -double -chlorid of di- methyl - diamido-phenaz- th ionium chlorid. B61. Nitromethylene Blue. 664. chlorid of dimethyldiamido- toluphenazthionium. TmOBKKZZKTL COLORS. 669. I ><'liydrothiotoluidin. quoroun colobs. •/uinophthalone. 670. Sulphohydro derivative of a polytnladn. 675. Tmocateohlrj S (composition unknown). DTDIOO < oi.oits. 512. Tetrabromofluorescein. 517. Tetraiodofluorescein. 520. Tetraiododichlorofluorescein. 521. Tetrabromotetrachloro fluor- escein. 527. Sodium bisulphite com- pound ofccrrulein. 504. Hydrochlorid of diethyl-m- amido-phenol-phthalein. 570. New Cray (composition un- known). .-.03. 000. Phenyl- and tolyl-safranins. Mixtures of diariilido-amido- trianilido-, and tetraani- lido - phenyl - phenazo - nium chlorids. Indulins and fluorindins. 692. IndiRotln disulphonio acid. 563. Dioxy-anthraquinone-b-qui- noiin. 584. Diamidophenyl and tolyl- tolazonium chlorids. 001. Sulphonated indulins. 050. Chlorid of tetramethyl-di- amido-phenazthionium. 680. ImliRotin. COMPILED DATA UXDER GREEX TABLE NUMBERS. 67 PHYSIOLOGICAL ACTION OF COAL-TAR DYES. SUMMARY OF SYMPTOMS. A rough summary of the symptoms noted or positively deter- mined to be absent, the number of the deaths produced, and the number of cases in which nothing abnormal was noticed may serve as a convenient guide in considering the detailed statements herein- after given relative to all the symptoms, clinical data, legislative and other publications, or permissions. In the following tables are brought together most, if not all, of the recorded observations with respect to the humans and other animals upon which the physiological action of coal-tar dyes has been studied as well as the results of autopsies when recorded. The columns headed " Unfavorable," " Favorable," and " Contradictory" have the same significance as in the preceding table; italicized numbers are those of the permitted colors of Food Inspection Decision No. 76. Asterisked numbers indicate that the dye was administered hy podermically . Experiments on Do«.- Table I. — Observations on dogs. (Reference to Green Table numbers.) Symptoms. Unfavorable. Favorable. Contradictory. Deaths 1; 2: 3; 488 or 490; 574; 639 (2); 649. 3; 488 or 490 ; 572; 57 16 (2); 55; 7(>v Autopsies 108; 105;300;467; K 188»; 480*; 801; I • Stained skin 639; 649. 1 Stained conjunctiva' Respiration difficulties Temperature normal Temperature hi^h Temperature low Genera] depression Weakness 1 1; a 2; 620; 649 407 461; 601. 488 or 490; 3 188 or 490 4.M. 426; t.49; 07.") 66a I; 3; fl 1'.; 488. Loss of weight 1: 488 W 190; 8 ' 17; 18; 4 • I; 65a Emaciation. . Loss of appetite ti50. Colore. 1 urine l; 502; 516; 861; 621; 1 L; 18; it: 16; I rine not affected l; 8; . 601. Albuminuria. . Albuminuria doubtful Alkaline urine 28; 166; 240 I L; 17, 18, or 11: 70; • Bladder Irritation Thirsl Thin stool Diarrhea Blood} stool. 68 COAL-TAR COLORS USED IX FOOD PRODUCTS. Table I. — Observations on dogs — Continued. Symptoms. Unfavorable. Favorable. Contradictory. Pus in stool ' 650 Stool desires 502 Vomiting Retching 1; 2; 3; 6; 97; 27" 488or490;51( 614; 639; 649 675 1 488 or 490; 639; 488 or 490; 639. ;42S >;532 670 649; ;479; ;574; 675. 675... 399; *•• 433; 467 576; 692... 11; 16; 86; 87; 95; 197; 269; 428; 448; 451; 483; 650. 16; 86. Salivation 584. Anemia Stupor 601. Desire to sleep 451. Catarrh of eves and nose . . 451. Cramps 1; 2 601. Convulsions 2; 675 Tremors 87. Congestion Paralysis of limbs 649 1; 675 16; 55; 87; 428; 450. Destroys coloring matter of blood. 1; 2 87; 601. 601. 650. i The following table discloses the results of the autopsies made and reported on dogs. Table II. — Autopsies on dogs. Symptoms. Unfavorable. Favorable. Contradictory. No change in internal organs. 87. 105; 399; 467; 477; 593 103 87 III; 457. No change except whole interior colored red. LTVER. Fatty degeneration 572; 620; 639 601; 650. 488 or 490... 451 (2); 480 II*; 601. Pale.. 451 (2). 574 649 650. Blood-Oiled . 574 KIDNEYS. 3; 639; 649 451 (2); 480* (II). Filled with decomposed blood corpuscles. 639 451 (2). 480 (1); 601; 650. Soft. 480* 11. 488 or 490 .. 601. Blood-filled 574 620; 639 692 1 IIC'l 650. 650. STOMACH. 488 or 490 650. Catarrh 188 oi 650. ; 801. with colored inw l.t \ Filled with blood :,: i 601. ■BAST. 574 Paraly* 16 (2); 87 I; 428; 87 II. COMPILED DATA UNDER GREEN TABLE NUMBERS. 69 Table II. — Autopsies on dogs — Continued. Symptoms. Unfavorable. Favorable. Contradictory. INTESTIN'ES. Colored 488 or 490 650. Catarrh 488 or 490 COLORED. Brain 650. Fat 650. Skin 650. Pleura Heart sac 650. Diaphragm 650. 488 or 490 650. 480*11. All organs swollen 480*11. General congestion Peritonitis 3(11) 70*. Fat all disappeared Flabby muscles G39 639 In this connection it may be of interest to note that out of 16 dyes producing death when administered through the mouth to dogs, 7 were on the United States market in the summer of 1907. Their Green Table numbers are as follows, the numbers in parentheses indicating the number of makers or importers, out of a possible 12: offering them: 55 (2) ; 86 (8) ; 95 (2) ; 428 (3) ; 451 (5) ; 601 (1) ; 650 (2). The following table gives the Green Table numbers of coal-tar dyes which on administration to dogs positively did not in certain specific cases produce the particular disturbances recited, although the case of the several colors as a whole is regarded in the literature as ■■favor- able," " unfavorable," or "contradictory," as stated. Table III. — Observations on dogs showing definite negative result* in specific grouped under the general verdict of the literature as a wholt . Symptoms. Unfavorable. Favorable. Contradictory. 014 4: 105; 462 K; 601. I >i irrlnM 03; 106: -\ ■ 521; 527; 576. ; Allmminiirhi 6; 97; 277; i 051; 659. 14; 16; 197; 801. 1 i92 That is, when these dyes were tried od dogs the observers reported in certain specified cases positively and definitely, the absence of any of the symptoms named. It is stated of the follow ing colors that they produced no bad effects in certain specified cases: \; 9; 55; B6j 105; L88; L97; 240; 157; 572; 593. No had effect except colored urine: 13, 95. Nb had effect except albuminuria: 88,287. N<> bad effect except loss of weight: 12.5 pei- cent (17. is, or 41). It is therefore t rue of all of these coal-tar d\ e> that then' are Condi- tions under which they have 0660 oh^erved not to produce the had 70 COAL-TAR COLORS USED IX FOOD PRODUCTS. effects as above set forth; but this tabulation must not be taken to mean that these dyes can not, under any conditions whatever, produce untoward results ; the reverse is true in most cases. Experiments on Human Beings. The Green Table numbers of those colors concerning which experi- mental data are available on humans are as follows : 1 ; 2 ; 3 ; 4; 6 ; 9; (17, 18,41); 55; 65; 85; 86; 95; 102; 103; 105; 106; 107; 197; 427; 428; 448; 462; 532; 602; 650. The symptoms produced are classified in the following list: Internally administered. Not poisonous 4; 9; 55; 65; 85; 95; 102; 103; 105; 106; 107; 448; 462 Poor general condition 86; 650 Fever 2 Loss of appetite 2 Vomiting 3; 532; 650 Intestinal irritant 650 Diarrhea 532; 650 Bladder irritant 650 Colored urine 86; 95; 650 Albuminuria 650 Increased micturition 650 Irritant 532 Inflammation 532 Bad taste in mouth 86 Restlessness 86 Rush of blood to head 86 Vertigo 86; 650 Headache 650 Delirium 650 Twitching of muscles 650 Yellow coloration of skin 3 Yellow-colored mucous membrane. 3 Food colored with it made a family sick 3 Adults withstand 1 Children and weak adults do not withstand 1 Deaths 2; 3 Autopsy l 3 Dryness of throat 86 It should be noted that of the 13 dyes here classed as not poisonous to humans all but No. 102 were on the United States market in the summer of 1907, as is shown in the following table: Number of dealers offering these nonpoisonous colors in 1907. Green Table numbers. Sources offering same. Green Table numbers. Sources offering same. Green Table numbers. Sources offering same. Green Table numbers. Sources offering same. 9 55 ,0 1 2 i 85 i| 103 id;, 106 (i 1 5 107 448 463 7 4 2 1 On permitted list, Food inspection Decision 76. Symptoms produced by external application of certain colors (Green Table nurnbett). I dermatitis 4 Eczema (17, 18, or 41); 197; 602 Inflammation 427; 428 Burning 2; 427; 428 Itching 2; 6; 427; 428 Blister* 2; 6 Swelling 427; 428 It should be noted thai No. 86 Ims been tried on humana and has been found aot to produce diarrhea <>r vomiting. 1 Bemorrnagic gastritis. COMPILED DATA UNDER GREEN TABLE NUMBERS. 71 G. T. 3 has apparently killed a human at 60 mg per kilo and the autopsy disclosed hemorrhagic gastritis. G. T. 448 has been sug- gested as a possible remedy for Blight's disease. It should also be borne in mind that adults can stand G. T. 1 in doses of 540 to 900 mg daily for a long time, whereas children and weak adults stand that substance only poorly. Experiments ox Small Animals. Results of experimenting on rabbits with 10 coal-tar dyes whose Green Table numbers are 1, 2, 86, 89, 107, 427, 448, 504, 517, and 563 have been tabulated as follows : Paralysis 427* Cramps 427* No harm produced. B9; 107; 44S; 504; 517 Death Diarrhea .... 1; 2; 427*; 563* 1 Colored urine 86 Softened feces 86 In the case of the numbers marked with an asterisk the color administered hypodermically. In this connection reference should also be had to the paper of Penzoldt abstracted in Section VIII, page 55. The four coal-tar dyes 55, 103, 425, and 480 have been tested on guinea pigs and no disturbance was noticed in all, but hi the case of 103 occasional thirstiness was observed. Xo. 448 has been fed to hens without damage, and Xo. 2 has been recommended as an insect- icide, a fungicide, and a mouse poison. GENERAL STATEMENTS. The following statements may be of interest before the detailed compilation is read: 1. O. Buss (Forschungsber. iiber Lebensmitttl, 1896, vol. 2, j>j>. 163-197, 237), in a paper entitled " Contributions to the Spectrum Analysis of some Toxic and Pharmacognostically Important Coloring Matters, with Special Consideration to the l'ltra-\ "inlet ." cites poisonous the following (Green Table numbers follow in parent wherever connection could be satisfactorily established): Picric Acid (1) Dinitroo- and p-Greeol (2) Martina Yell m (3) Aurantia (6) Fast Yell, m (8) Orange II (86) As nonpoisonous: Naphtfaol Yellow (4) Eoain Erythremia Anilin Blue As doubtful: Metaail Yellow Oorallin Stiraniii Methylene Blue [odin Green AUsarin Bl Naphthol Green Malachite Green Dahlia (450, Methyl Violet (451,464) | Veeuvin 72 COAL-TAR COLORS USED IN FOOD PRODUCTS. On the following Buss is noncommittal: Auramin (4-5) Biebrich Scarlet (163) Water blue (480) Magenta (448) Aurin J (483) Acid Green (435) Naphthol Black of various brands, which seemed to be mixtures. 2. It has been pointed out that the following nine colors are harm- less: Naphthol Yellow (4) Naphthol Brown (?) Chinolin Yellow (667) Pyrotin RRO (115) Acid Green (434, 435) Wine Green (?) Milling Red (?) Azo-acid Blue (36) Fastrose (?) Most of these have not been examined experimentally, but scien- tific studies have been made of the poisonous qualities of Azo-Blue and Naphthol Yellow. (Zts. angew. Chemie, 1896, p. 21^.) 3. Chlopin in Ins monograph (see p. 75) says: On the basis of my personal experience I consider the testing of the action of coal- tar dyes on man not permissible, since such experiments may induce in the subjects of experiment more or less serious symptoms of poisoning, for which in some cases there are no antidotes at our disposal. To such accidents, in my opinion, only the experimenter himself may subject himself, because he knows what he is doing. Pre- liminary tests of dyes on dogs and other animals afford no guaranty of escape from dis- agreeable accidents which may take place in the testing of the dye on man. {Page 111.) These data and calculations convincingly prove how erroneous the current opinion is that for the coloring of food substances and beverages only exceedingly small, almost unweighable, quantities of dyes are used. {Page 113.) On page 221 et seq., the following general discussion of tins subject is found: The Manner of Action of Poisonous Dyes on the Animal Organism. The mechanism and the chemistry of the action of the artificial dyes of the aromatic series on the animal organism remains to the present day, with few exceptions, exceedingly slightly and superficially studied. The same can be said also concern- ing the pathological and anatomical changes which are induced by these dyes. More than the others there have been studied in the toxicological respect the dyes belonging to the Nitro group; Picric Acid; Martius Yellow (Dinitro-naphthol potas- sium), and Saffron substitute (Dinitrocreeol potassium.) According to Kobert, these dyee belong to the poisons acting on the blood. Accord- ing to tli'i same authority, Methylene Blue, which belongs to the other chemical group of Thiazins, acts similarly . In the fundamental works od the sanitary investigation! for the dyes, by Cazeneuve and Lupine, by Weyl, and by Bantori, We find almost, no material relating to the explanation of the manner <>i tin' action oi the dyes. These Investigators limiting themselves to a very cursory description of the symptoms of poisoning, and reciting in most general terms the results <>f autopsy, not ei en indicating the cause oi death. Such, for instance, are the reporti of autopsies made by T. Weyl and wmo other invest Igator , a - quoted aboi e. J Apparmtly B purified formol Corallin. COMPILED DATA UNDER GKEEX TABLE XUMBEES. 73 We may expect that more detailed investigations in this respect will be made at the proper time by pharmacologists, since study of the mechanism and chemistry of the action of the poisonous substances on the animal organism is their province; for the hygienist it is quite sufficient merely to establish the fact that a given substance is poisonous or harmful, and he need not go any further. For this reason, in those cases in which I desired to clear up the causes of death of the animals in my experiments, and to record pathological and anatomical changes (although by the terms of the regu- lations governing this competition, a close study of the action of the dyes, and the ascertainment of the mechanism and chemistry of their action was not required), I called in a person more competent than myself on these questions. Not counting the duplicates we made five autopsies all told. In all these cases death resulted from paralysis of the heart. The pathological and anatomical changes in all cases, except one, did not present anything specific, and finally reduced themsel a feebly expressed turbid swelling of the heart and of the liver, a rush of blood to the stomach, and a congestion of the internal organs. The exception was the autopsy of a dog, which died from Methyl Orange; this dog died with the symptoms of paralysis of a cerebro-spinal nature. This experiment was made twice, and the autopsies of both animals showed hyperemia in the lowest part of the spinal column, on the border of the anterior and the lateral columns. As to the symptoms of poisoning not resulting in acute death, here most frequently was observed vomiting, diarrhea, and albumen in the urine, showing disturbance of the functions of the digestive tract, and an affection of the kidneys. A highly typical picture of poisoning is presented by the sulphid Vidal dyes. They cause rapid, almost instantaneous, deafening of the animal, whereupon the animal falls on one side in convulsions and lies, not moving its body, but convulsively and rapidly twitching its anterior limbs during several minutes. The tongue hangs out of the mouth, a strong secretion of saliva is noted, then vomiting begins, and the dog gradu- ally begins to revive; with difficulty he arises on his front feet and sits down, not being able yet to stand on his posterior extremity, which is in a state of paresis. After a few hours the dog becomes normal. The symptoms of poisoning, just described, are exceed- ingly similar to the supposed "apoplectic form" of poisoning by hydrogen sulphid, which had been studied on animals by K. Lehmann, and which was observed in per- sons who inhaled air containing a few per cent of this gas. Air containing 0.1 to 0.3 per cent of hydrogen sulphid kills cats and dogs in 10 minute-. In OUT experiments in which were introduced substances containing sodium sulphid, the poisoning must have been caused by hydrogen sulphid which was liberated from the dye by the acid of the gastric juice, and which could cause poisoning also through the stomach and through the respiratory apparatus. Fortunately Vidal dyes, owing to their repulsive odor, will scarcely find a wide application in coloring food and beveragi Some Reflections R eg a b him. IVihiii i: in VK8TIGATXONS OF Dybs FROM ▲ Sani iaky S iw DPOINT. The present investigation, as well ai all the investigations oi the action of dyes on animal- by previous investigators, had for its object the solution of the question in what number there exists among the dyei of the aromatic .- -v\<-< dyes which possess pwi-.ip.u-, or more or l< ■ pronounced harmful properties (answering essentially the sanitary toxicologi< al question). Prom the practical point of view such investigations presented and do present the most important interest, inasmuch as they afford a possi- bility of protecting the public from the use oi obvious) 11 and harmful sub- stances, but by such investigations question! iingly sanitary importance are not answered, namely: 1. Ought we to consider as quite harmless those dyes vrhich do not induce pro- nounced symptoms o! poisoning and which are designated herein l>v the term nan- poisonous? 74 COAL-TAR COLORS USED IX FOOD PRODUCTS. 2. Is the usual answer which is given to the hygienist by the defenders of the unre- stricted use of the coal-tar dyes for coloring food products and beverages, namely, that in practice the dyes are introduced into the human organism in so small quantities that their properties can be neglected, justified? To both questions, besides the facts and considerations which I gave above in my investigations, we may reply experimentally in two ways: (a) By prolonged investigations continued over a period of years of the action of very minute quantities on the animal organism, which has so far, owing to the incon- veniences of such long experiments, not been done by anybody; and (6) By investigation of the action of small doses of dye on some physiological func- tions, and first of all on the activity of the digestive organs, which is first of all dis- turbed upon the introduction of dyes into foods and beverages. The solution of the last question can be best promoted, in my opinion, by experiments made on dogs and by exploratory examination of the body according to the method of Prof. Pawlow. Unfortunately, experiments such as these, owing to the difficulty of the Heidenhain- Pawlow operation, are inaccessible to the majority of investigators. As a very useful substitute of such investigations may serve observations on the action of dyes on the activity of the digestive juices outside the body of animals. On my proposition Dr. A. E. Winogradow began in my laboratory experiments on the action of small doses of dyes of the aromatic series on the digestion in vitro. Dr. Winogradow so far examined 25 coal-tar dyes in this respect according to the method of Metta and convincingly proved that in insignificant doses coal-tar dyes (from one-half to 4 milligrams) entirely stopped the digestion of albumen by artificial gastric juice. It was found that the capacity to depress the digestion is possessed not only by poison- ous dyes, but also by dyes which proved in my experiments on animals nonpoisonous. It is quite possible, therefore, that an admixture of coal-tar dyes will exert an unfa- vorable influence on the digestion and assimilation of food prepared from products col- ored by them. Experimental proof of the last supposition can be given only by experi- ments on the influence of dyes on the metabolism of substances in animals and man. COMPLETE DETAILED STATEMENT OF ALL COMBINED DATA. ABBREVIATIONS OF AUTHORITIES CITED. The data hereinafter given is brought together as nearly as pos- sible under the Green Table numbers to which it is pertinent. It is believed that the literature has been quite thoroughly searched, and that nothing of substantial importance has escaped recording in this compilation; certainly whatever may have escaped can hardly serve to change the general conclusion to which this compilation leads. In order to avoid repetition in the following tabulation, "Weyl" is to be understood as referring to the book entitled "The Coal Tar Colors, with Especial Reference to their Injurious Qualities, etc.," by Theodor Weyl, translated by Leffmann and published in Philadel- phia, Pa., in L892. "Iieber" refers to the book entitled "The Use of Coal Tar Colors in Food Products," by Ilu.L r <> Ldeber, published in New York in 1904. "Fraenkel" refers to the book entitled "Arzneimittel Synthese," by Dr. Sigmund Fraenkel, published in Berlin in 1900. "Confectioners List" refers to the Official Circular from the Exec- utive Committee of tln v National Confectioners' Association of the COMPILED DATA UNDER GREEN TABLE NUMBERS. 75 United States entitled " Colors in Confectionery" and reprinted, in part, in the book entitled "Food Inspection and Analysis," by Albert E. Leach, published in Xew York in 1906, pages 630-634. " Resolutions of Swiss Analytical Chemists" refers to these reso- lutions as published in Zeitschrift fur Xahrungsmittel Untersuchung und Hygiene, 1891, page 293. "Schacherl" refers to SchacherFs publication entitled "Die Zulaes- sigkeit Kuenstlicher Farbstoffe zum Farben von Lebensmitteln," published in Vol. Ill, pages 1041-1048, of the Report of the Fifth International Congress of Applied Chemistry held in Berlin June 2 to 8, 1903. "Chlopin" refers to Chlopin's monograph published in Russian and entitled "Coal Tar Dyes. Classification, properties, and action of artificial dyes on the animal organism, etc.," published at Dorpat in 1903, or to the abstract of Chlopin's paper printed at page 169-172 of Vol. IV of the Report of the Fifth International Congress of Ap- plied Chemistry held in Berlin, 1903. "Canton of Tessin" refers to the publications of the Tessin regula- tions published in 1897 in Zeitschrift fur Untersuchung dor Nahrungs und Genussmittel, page 414. Whenever possible the doses administered have been given in mil- ligrams per kilo and grains per 100 pounds of body weight of animal. In the case of the tabulations taken from Chlopin's monograph this was not done; but in order to render such comparative data easily available factors have been placed at the head of each tabulation; for example, under G. T. 6 (1 gram =106 mg = 74.2 grains), which means that each gram administered amounts to 106 mg per kilo or 74.2 grains per 100 pounds of body weight of animal; by multiplying the doses given by either of the factors the corresponding compara- tive information is obtained. In addition to the 106 Green Table numbers that have been examined physiologically, there are reported the results of the phys- iological examination of 8 coal-tar colors mot in the Green Tables, of which 3 are said to be nonpoisonous, .'5 are called poisonous, 1 is called harmful, and the last is said to be "nol quite harmless.' J These 8 dyes are qoI included in this compilation. TABULATION Bl GREEN TABLE NUMBERS <>! PHTSIOLOGICA] \.\i> 01 mi: D \ i \. G. T. 1. Tradi mums. — Picric acid; carbazotic acid. Scientific narru .—Symmetrical trinitrophenol. Discovered. — 1771. Shade, — Yellow. Not offered. 76 COAL-TAR COLORS USED IN FOOD PRODUCTS. FAVORABLE. Nothing. UNFAVORABLE. 1. Prohibited by Confectioners' List. 2. Weyl (p. 30): "The injurious character of picric acid has long been known." 3. "In Germany its employment for coloring food is forbidden by the imperial enact- ment of 1888, on account of its poisonous character." (pp. 68-71.) 4. "The foregoing statements show that while the acid must be considered poisonous, its injurious character is far less than has generally been assumed, nevertheless, the legal prohibition of its use as a coloring matter for food or drink is just." (p. 71.) 5. "Erb gave a rabbit weighing 1,700 grams, 0.06 gram of potassium picrate (24.5 grains per 100 pounds) daily for 90 days; slight loss of weight and occasional diarrhea were noted, but nothing more serious." (p. 69.) 6. A rabbit weighing 2,065 grams died at the end of 19 days, after having taken 2.52 grams of the substance, or 854 grains per 100 pounds body weight; number of doses not stated, (p. 69.) 7. Weyl's experiment on a dog, weight not given: April 21-26, 0.24 gram (3.7 grains) sodium picrate daily; April 28-May 9, 0.36 gram (5.5 grains) daily; total, 5.76 grams (8.9 grains) sodium picrate; no serious disturbance; May 13, 1.2 grams (18h grains) sodium picrate at one dose; weakness marked, diarrhea and dyspnea next day; May 14, 0.6 gram (9^ grains) caused vomiting; evening of same day, 0.36 gram (5.6 grains) given; May 15, animal lively; 0.24 gram (3.7 grains) again given, and on evening of same day 0.72 gram (11.2 grains); May 16, marked weakness of animal, and 0.16 gram (2.5 grains) given, causing vomiting; May 17, 0.17 gram (2.6 grains) given; May 18 and 19, animal definitely recovered, and aside from strong yellow tinge of the conjunctiva and skin, no abnormal conditions manifest. Animal died May 20 after receiving 1.32 grams (20.4 grains) potassium picrate. Weyl concludes, therefore, that dogs are resistant to this substance, notwithstanding the prostration and the blood disorganiza- tion. 8. Weyl summarizes the effect on humans from therapeutic and poisoning cases to the effect that daily doses of from 0.54 to 0.90 gram (8.3 to 13.8 grains) of potas- sium picrate are easily borne by healthy adults for a considerable time; children and weak adults bear picric acid badly, (p. 70.) 9. "Picric acid * * * is poisonous * * *." (p. 96.) 10. Fraenkel (p. 572): "On the other hand, this substance is not usable for interna] administration on account of its decomposing the red blood corpuscles, and of it energetic cramp production, as well as on account of its disturbance of the kidneys, and the ultimate paralysis of the respiratory centers; nevertheless, picric acid is not to be considered a violent poison * * *." 11 Schacherl (p. 1044): "Picric acid * * * (is), according to numerous state- ments in the literature, poisonous even in small doses, and (is) therefore un- qualifiedly to be declared as impermissible." L2. LlBBEB (p. 16), where it is stated to be forbidden by the German law, aiul is also otherwise substantially the same as Weyl above quote I. i.; Resolutions "i" tin- Society of Swiss Analytical Chemists, September, L891: "The following are to be regarded a- < < » 1 < . r i 1 1 lt matters harmful to health: * * * picric add * * *." n. Prohibited by the Belgian law of June 17, L891. ii l.i .win (Lekrbuch der Toxtkologie t t897 % p. lSt)\ "Picric acid is poisonous. Rab- bits Can stand daily I" milligrams of a green containing picric acid, but n<>( L'O milligrams. Their death Is accompanied by paralysis." It;. JJissli.-l ii ,i ■ poisonous. COMPILED DATA UNDER GREEN TABLE NUMBERS. 77 G. T. 2. Trade names. — Victoria Yellow; Victoria Orange; Golden Yellow; Saffron Substitute; Anilin Orange; Di-nitro-Cresol. Scientific name. — Di-nitro-o-and-p-cresol. Shade. — Yellow. Xot offered. Discovered. — 1869. Used for coloring butter, liqueurs, etc. FAVORABLE. Nothing. UNFAVORABLE. 1. Prohibited by Confectioners' List. 2. Fraenkel (p. 572): "On the other hand dinitro-cresol is much more intensely poisonous (than picric acid), which is probably caused by its greater solubility in water." 3. Schacherl (p. 1044): * * * Dinitro-cresol [is], according to numerous state- ments in the literature, poisonous even in small doses, and [is] therefore unquali- fiedly to be declared as unpermissible." 4. Resolutions of the Society of Swiss Analytical Chemists, September, 1891: "The following are to be regarded as coloring matters harmful to health * * * Dinitro-cresol * * *." 5. Forbidden by the Canton of Tessin. 6. Weyl (p. 31): "I have shown the same (poisonous nature) for Dinitro-cresol (Saf- fron Substitute). (See Zts. angew. Chem., 1888, No. 12, for confirmation of my results by Gerlach.)" 7. "The reverse is the case with the poisonous dinitro-cresol (Saffron Substitute). " (P- 55.) 8. Weyl describes experiments with this compound, {pp. 71-85.) 9. Fourteen rabbits were experimented on, of which 13 died. Amounts administered in the fatal cases per 100 pounds body weight were {p. 74): Grains. Grains. 189 175 175 168 175 175 175 168 175 168 175 175 175 Of 12 experiments on dogs, 5 receiving the color by the mouth and 7 hypoder- mically, 3 cases resulted fatally; the fatal case by the mouth requiring ;> s '. grains 1 >cr LOO pounds body freight ; the 2 fatal cases hypodermically re p r es en ted 1 1 and 20 grains per LOO pounds body weight . reaped ively, although L40, 88.5, :;i .6, and •-..in- per i<><> pounds body weight by the moutb were borne without fatal effect; ami lm^, m ad 4.9 grains per loo pounds body weight, hypo- dermically, wen- also borne \\ ithout fatal effect {p. 75). lo. \Vi.yi.i'/>. 9€)'. "* * * Dinitro-c re sol * * * are [is] poisonous; * * *" ii. Prohibited by the Belgian law of June it, L801. Ztt. Nahr.G Eti ommended as an Inse cti cide, L500 being sufficient far all ordinary purposes. One milligram is sufficient to kill a mouse; 2 milligrams recommended f'>r killim: mice. ]:;. W'kvi. ( HandbuA der Hygieru i: For humans, thefataldose, when administered by the stomach, appears t<> be no milligrams per kil<> body weight, or 48 graii 100 pounds. 78 COAL-TAR COLORS USED IN FOOD PRODUCTS. 14. "The president of the Council of Oppeln forbids on April 19, 1899, the use of Saf- fron Surrogate for coloring food products." 15. Lewin t (Lehrbuch der Toxikologie, 1897, p. 232): "Saffron Surrogate * * *, which is used for coloring foodstuffs, is poisonous. It appears to attack the coloring matter of the blood, and produces, in the case of dogs, vomiting, cramps, and convulsions. * * * Feathers colored with Saffron Surrogate cause burning and itching, and finally blisters on the hands of the women working with them; the faces were also similarly affected, and this was accompanied by loss of appetite and fever." 16. Buss lists it as poisonous. G. T. 3. Trade names. — Martius Yellow; Naphthol Yellow; Naphthylene Yel- low; Naphthylamin Yellow; Manchester Yellow; Golden Yellow; Saf- fron Yellow; Jaune d'Or; Jaune Naphthol. Scientific name. — Dinitro-alpha-naphthol. Discovered. — 1864. Shade. — Yellow. Not offered. FAVORABLE. Nothing. UNFAVORABLE. 1. Prohibited by Confectioners' List. 2. Weyl (p. 31): "Cazeneuve and Lepine pointed out the poisonous nature of Martius Yellow * * *." 3. "This body (Chamber of Commerce at Sonneberg) recommends for the prepara- tion of children's toys three colors, the poisonous character of which I can demonstrate. These are Martius Yellow * * *." (p. 34.) 4. "For instance, for preliminary researches, dogs and rabbits have value for chemical reasons. The conclusions derived from such experiments must be accepted with great deliberation, since it happens that rabbits will bear without injury doses which will seriously, nay, even fatally, act upon the dog, as I have already shown to be the case with Martius Yellow." (p. 56.) 5. Where two experiments by Cazeneuve and Lepine are referred to, in which diarrhea, vomiting, and albuminuria were produced by this substance. (pp. 85-89.) 6. Weyl's own experiments on 4 dogs showed weakness, vomiting, diarrhea, and albuminuria resulting from the use of this color; the amounts of color admin- istered per kilogram of body weight were 73, 17.5, 17.5, and 11.3 milligrams, which amount to 51, 12, 12, and 8 grains, respectively, per 100 pounds of body weight, (p. 87.) 7. "Martius Yellow, therefore, belongs to the injurious colors. As a coloring matter for food and drink its use should be wholly prohibited." (p. 89.) 8. "* * * and Martius Yellow are poisonous; * * *." (p. 96.) 'i l.cwiv ( Lehrbuch der Toxikologie, t897 t j>. tSl): "Like Saffron Surrogate, it is poisonous. In an experiment on myself, using large doses, I noticed among others Ihe general yellow coloration of the skin. In a poisoning resulting fatally after 5 hours with Martins Yellow, vomiting, yellow coloration of the skin and mucous membranes were observed; whereas the autopsy revealed, among other things, hemorrhagic gastritis, (Jacobson, Hosp. Tid., 18W, p. COMPILED DATA UNDER GREEN TABLE NUMBERS. 79 10. " Such small amounts as are used for the coloring of pastry are said to be non- poisonous. ( Vitalil boll, chim.farm., 1893, p. 738.)" (p. 231.) 11. Cazeneuve and Lepine (Compt. rend., 1885, v. 101, pp. 1167-1169) say: I. "A dog received 71 milligrams per kilogram of body weight, or 50 grains per 100 pounds daily. On the second day diarrhea and vomiting ensued; loss of appetite except for milk. Thereafter it experienced difficulties in breathing; suffered albuminuria; its urine was colored; it died on the sixth day. The autopsy disclosed considerable congestion. II. A dog weighing 22 kilo- received 400 milligrams (19 milligrams per kilo or 13 grains per 100 pounds); this caused a yellow vomit; next day it received 500 milligrams (27 milligrams per kilo or 17 grains per 100 pounds), which caused violent diarrhea, fever, thirst, disinclination for all food. The animal was killed; the autopsy showed badly congested kidneys." 12. Prohibited by the Belgian law of June 17, 1891. 13. Schacherl (p. 1044)'- "* * * Martius Yellow * * * [is], according to numerous statements in the literature, poisonous even in small doses, and [is] therefore unqualifiedly to be declared as unpermissible." 14. Prohibited by law in Italy. (See Lieber, p. 24.) 15. Fraenkel (p. 572): "This substance also shows poisonous properties, although it is less poisonous than dinitro-cresol." 10. Resolutions of the Society of Suiss Analytical Chemists, September, 1891: " The following are to be regarded as coloring matters harmful to health * * * Martius Yellow * * *." 17. Forbidden by the Canton of Tessin. 18. Dietrich (Zts. Nahr. Genussm., 1902, v. 5, p. 364)'- "A lot of groats, after eating which a family became sick, was found to be free from ordinary poisons, but had been colored with Martius Yellow." 19. Buss lists it as poisonous. DOUBTFUL. 1. Winogradow (Zts. Nahrs. Genussm., 1903, v. 6, p. 589) says it noticeably retards digestive action; is not indifferent. G. T. 4. Trade names. — Sulphur Yellow; Sulphonaphthol Acid Yellow: Succinic; Solid Yellow; Saffron Yellow: Jaune Acide (\: Jaune Acide; Fasl Yellow; Citronin; Anilin Yellow ; Acid Yellow S. Names und< r which it was offered on 'tin United States market as w. Offered by 10 out of L2 soun I A\ <>l; Alii | I. Permitted by ( kmfecl Loners' !.i.-t . \/i\m\i wi. I mim {Compt rend., Wl, pp. 1167-1169): ■ A dog day for 5 days 32 milligrams per kilogram of body weight, «>r 2:\ grains per n><> pounds; for the i<> days next rfu< i ived four times that amount, that i-. L33 milligrams per kilogram ol body \\<'i'.:lit. or 93 grains 80 COAL-TAR COLORS USED IX FOOD PRODUCTS. 2. Cazeneuve and Lepine (Compt. rend., 1885, v. 101, pp. 1167-1169) — Continued. per 100 pounds; for the 10 days next succeeding it received daily twice the last amount, or 266 milligrams per kilogram of body weight, that is, 186 grains per 100 pounds. It received altogether in the 25 days 62£ grams, or 964 grains. There was no vomiting, no diarrhea, and no albumen in the urine at any time." 3. Weyl (p. 31): "* * * not poisonous to human beings and dogs: Naphthol YellowS. * * *." Weyl describes his own experiments on 3 dogs, giving them, respectively, 417, 34, and 100 milligrams per kilo body weight, or per 100 pounds 292, 24, and 70 grains, respectively. Whether the color was administered by the mouth, or injected subcutaneously, all bodily functions appeared to remain normal, and it was only in the case of repeated doses of 417 milligrams per kilogram of body weight, or 292 grains per 100 pounds, that albuminuria appeared. {pp. 89-92.) 4. "Only the sulphonated colors Naphthol Yellow * * * are harmless and appli- cable to the coloring of food and drink." (p. 96.) 5. Permitted by the laws of Austria. 6. Permitted by the law of Italy. 7. Permitted by the law of France. 8. Schacherl (p. 1044)' "* * * Naphthol Yellow S * * * possesses no poi- sonous properties." 9. Fraenkel (p. 572): "Naphthol Yellow S is an entirely nonpoisonous substance." 10. Meyer (/. Amer. Chem. Soc. 1907, v. 29, p. 900): One hundred milligrams per kilogram of body weight for the initial administration, and subsequent admin- istrations increased geometrically. After the second administration intermit- tent diarrhea resulted, emphasized by increased amounts with no albumin or sugar in the urine; continued for 14 administrations; 60 that in 14 administra- tions 147.58 grams of color had been given; the initial dose is 70 grains per 100 pounds of body weight, and the average daily dose of the total administered is 394 grains per 100 pounds of body weight. Urine only slightly yellow col- ored after small doses, but red after larger doses. 11. Lieber (p. 14S): A dog received 36 milligrams per kilogram of body weight, or 25.2 grains per 100 pounds once a day seven times every other day; during the whole period the dog was apparently in good condition with no bad effects from the color. 12. Buss lists it as nonpoisonous. 13. Cazeneuve and Lepine (Compt. rend., 1885, v. 101, pp. 1167-1169): Three chronic invalids received daily from 2 to 4 grams of the dye in cocheta; except slight colic and diarrhea nothing abnormal. DOUBTFUL. 1. Lewin (Lehrbuch der Toxikologie, 1897, p. 282): "Acid Yellow S is said to be able t'» produce < 1«tiiki j it i ~ -Hi frequent contact therewith." 2. WlNOORADOW {Zls. Nulir. Qenuum., 1908, v. 6, p. 589) says it noticeably retards digestive action; ianol indifferent. G. T. 5. Trade name. — Brilliant Yellow; Naphthol Yellow S or RS. Scientificnantc. — Dinit m-alplia-naphthol-alpha-monosulphonicacid. Discovered and patented. — 1884. Shade. — Yellow. Not offered. COMPILED DATA UNDER GREEN TABLE NUMBERS. 81 FAVORABLE. 1. Permitted by Confectioners' List. 2. Weyl (pp. 92-94)'- Describes experiments on 2 dogs, in which the initial dose was 532 milligrams by the mouth, and 17 milligrams hypodermieally, per kilo- gram of body weight, respectively, 372 and 12 grains per 100 pounds of body weight; in both cases the urine was colored, in the second case traces of albu- minuria resulted. In the first case the albuminuria was doubtful. 3. "Brilliant Yellow is not poisonous, even in large doses, when administered by the stomach * * * the albuminuria was very slight, (p. 94-) 4. "Only the sulphonated colors * * * Brilliant Yellow, are harmless, and applicable to the coloring of food and drink. " (p. 96.) 5. Schacherl (p. 1044)'- " * * * Brilliant Yellow * * * possesses no poisonous properties. " 6. Fraenkel (p. 572): "For the same reason * is without effect. " G. T. 6. * Brilliant Yellow Tradenames. — Aurantia; Nitrodiphenylamin; Imperial Yellow Kaiser Yellow. Scientific name. — Hexanitro-diphenylaniin. Discovered. — 1873. Shade. — Yellow. Not offered. Nothing. FAVORABLE. UNFAVORABLE. * Aurantia suspicious. " * * * Aurantia [is] according to numerous statements 1. Weyl (p. 00): " * 2. Schacherl (p. 1044) in the literature poisonous even in small doses, and [is] therefore unqualifiedly to be declared as impermissible. " 3. Chlopin {p. 1 16) as results of his experiments considere it injurious. The experi- mental . Forbidden by the Canton of Tessin. 7. Lbwin (Lehrbuch der Toxihologie, 1897, p. 232): "Aurantia has a poisonous action. Alter wearing gloves for 8 hours made with so-called dogskin, which were col- ored with Aurantia, a man suffered confluent blisters, accompanied by itching. The workmen with this material get blisters on the face and on the hands. Perspiration increases the tendency to such blisters." 8. Buss lists it as poisonous. G. T. 8. Trade names. — Acid Yellow; Fast Yellow G; Acid Yellow G; Fast Yellow; Fast Yellow extra; Jaime Acide; New Yellow L. Names under which it was offered on the United States market as a food color in 1907.— Fast Yellow Y; Fast Yellow G; Acid Yellow G pat.; Fast Yellow 053. Scientific name. — Amidoazobenzene-disulphonate with some sodium amidoazobenzene-monosulphonate. Discovered. — 1878. Shade. — Yellow. Offered by 5 out of 12 sources. FAVORABLE. 1. Permitted by Confectioners' List. 2. Cazeneuve and Lepine (Bull, de Vacad. de med., April 27, 1886, p. 643), where it is classified among the "nontoxic " colors. :;. Fraenkel (p. 575), where it is stated to be nonpoisonous. 4. Permitted by the law of Austria. UNFAVORABLE. 1 . Weyl (p. 115): "Poisonous to human beings. (?)" 2. Chlopin (p. 151) considers that the work of others makes this a suspicious color. 1 1 is own experimental data are as follows: Experimental data by Chlopin. <;. T. 8 AND 9. [i grain -152 mg— 106 grains.] Weight. 24 hours' urine. ( fenei \i oondltton of animal and mine. 1901. LI 13 Grams. Kih.s. cc. 310 Dog normal; color urine normal; add; no albumen. it 1 6.6 Do. IS Do, Lfl Urine greenish brown; reddens with HsSO< and HQ; no 17 880 820 albumen. Color same; traoea of albumen Is Color leas Intense; trace of albumen. I'.t Do. Urine greenish brown; albumen gone; dog is lively. 21 202 Everything normal. Total 4 ( oncluaion; "Suspicious," COMPILED DATA UNDER GREEN TABLE NUMBERS. 83 3. Buss (Forschungsber. iiber Lebensmittel, 1896, Vol. Ill, p. 173): Is regarded as poisonous. 4. Robert (Lehrbuch der Intoxicationen, 1893, p. 335): Listed as poisonous. 5. Lewin (Lehrbuch der Toxikologie, 1897, p. 231) says "produces eczema," and cites Deutsche Med. Wochenschr., 1891, p. 45. G. T. 9. Trade names. — Fast Yellow R; Fast Yellow; Yellow W. Name under which it was offered on the United States market as a food color in 1907.— Fast Yellow 034. Scientific name. — Sodium salt of amidoazotoluene-disulphonic acid. Discovered and patented. — 1878. Shade. — Yellow. Offered by 1 out of 12 sources. FAVORABLE. 1. Permitted by Confectioners' List. 2. Weyl (p. 31): u * * * not poisonous to human beings and dogs * * * Solid Yellow." 3. Cazeneuve and Lepine (Compt. rend., 1885, v. 101, pp. 1167-1169): A. A dog received 42 milligrams per kilogram of body weight, or 29 grains per 100 pounds, for 5 days; thereupon received four times that amount for 5 days, or 168 milli- grams per kilogram body weight, or 117 grains per 100 pounds; for the 10 days next succeeding it received daily twice the last dose, or 336 milligrams per kilogram of body weight, that is, 235 grains per 100 pounds; it then received in 1 day 20 times the original dose, or 840 milligrams per kilogram body weight, that is, 596 grains per 100 pounds, and during the entire period nothing abnormal was noticed. B. Three chronic invalids received from 2 to 4 grains of the dye daily; except colic without diarrhea nothing abnormal. They con- cluded that this dye is no more harmful than Naphthol Yellow S (G. T. I . 4. Cazeneuve and Lepine (Bull, de Vacad. dc mcd., 1886, p. 643): Tolerated by man, well or sick. I \ FAVORABLE. 1. Weyl (p. 115): "Poisonous to human beings. (?)" 2. Chlopin (p. 151): Where he considers that the work of others makes this a sus- picious color. For his experimental data thereon sec table under G. T. 8; Chlopin's chemical description of the dye used applies to both Q. T. 8 and 9. :;. Cobibt Lekrbv/ch der Intoxicationen, t893,p.SS€): Listed as poisonous. G. T. 11. Tradi names. — Sudan I ; Carminaph. Na/ms a a J, /• which it was offered on Qu United States market as a food <<>h>r in 1907. —oil Orange 7078j Geraaio Orange I. Scientific name. — Benzene-azo-betanaphthol. Pi scon mi. L883. Shade, Orange Yellow. Offered by 2 out of 11* sources. i LVOB mii I. 1. Wan "Nonpoisonous * * * Soudan I * * *." 2. " Other Aso-colors, * * * for instance Soudan I • * ■ are entirely non- poisonous," '/>. / 84 COAL-TAR COLORS USED IN FOOD PRODUCTS. UNFAVORABLE. 1. Prohibited by Confectioners' List. 2. Weyl (p. 119): Dog received 18 grams in 6 doses in 17 days, and remained under observation 5 days longer; 4 doses of 168 milligrams per kilogram body weight (118 grains per 100 pounds) and 2 doses of 420 milligrams per kilogram body weight (294 grains per 100 pounds) produced colored urine, phenol in urine, vomiting and distinct albuminuria beginning with the third dose. From the foregoing Weyl concludes as follows: "The color in the doses administered is not entirely harmless, since a limited albuminuria seems to be brought about." 3. Fraexkel (p. 576): u It is not wholly harmless, since this coloring matter seems to produce a slight albuminuria.' 1 G. T. 13. Trade names. — Ponceau 4 GB; Crocein Orange; Brilliant Orange; Orange GEX. Names under which it was offered on the United States market as a food color in 1907. — Crocein Orange 10234; Crocein Orange; Crocein Orange G; Ponceau 4 GB. Scientific name. — Anilin-azo-betanaphthol-monosulphonic acid. Discovered. — 1878. Shade. — Orange Yellow. Offered by 6 out of 12 sources. FAVORABLE. 1. Permitted by Confectioners' List. 2. Weyl (p. 115): "Nonpoisonous * * * Ponceau 4 G B * * *.*' 3. Experiment on dog in doses of 161 milligrams per kilogram body weight; that is, 113 grains per 100 pounds body weight, apparently produced no disturbance aside from coloring of the urine, (p. 124.) 4. Weyl (p. 148): "Other Azo-colors * * * for instance * * * New Coc- cin * * * are entirely nonpoisonous. " 5. Weyl's conclusion reads as follows: "This color can be regarded as nonpoisonous." 6. Fraexkel (p. 577): "Ponceau 4GB can be regarded as nonpoisonous." UNFAVORABLE. 1. Excluded by law of Ac stria. G. T. 14. Trade names. — Orange G; Orange G G. Names under which it was offered on the United Stat. 7 Five hours after administration vomiting and involuntary thin feces. Dog does not eat; drinks much; weak; urine dark yellow; no vomiting. Do. Dog lies in cage; moves slowly when allowed out of cage; in the evening retching. Dog is still weak, but general condition somewhat better; urine strikingly colored; acid; no albumen. Began to eat; no albumen. General condition improving and almost normal; lassitude continues more than usual; urme normal color; acid; do albumen. In the evening repeated vomiting; lossofapetite. No vomiting; in the evening bothersome; does not eat. Stools normal; striking disturbance of movements; paralysis of legs; must spread hind legs to stand; when pushed not only falls, "but turns on his back; sight and hearing are normal; dog died at night. 1 I his dog ran away - Qrueblor's make. Berlin make. G. T. 17, 18, 41. 17. Trade names. — Chrysoidin Y; Clirvsonliii crystals. Name under which it was offered on the United States market as a food color in 1907,— Ghrysoidiu V. Scientific name. — Anilin-azo-meta-phenylene-diamin. Discovered. —1875. COMPILED DATA UNDER GREEX TABLE NUMBERS. 87 Shade. — Orange. Offered by 2 out of 12 sources. 18. Trade names. — Chrysoidin R; Cerotin Orange; C extra; Gold Orange for cotton. Name under which it was offered on the United States market as a food color in 1907. — Chrysoidin R. Scientific name. — Anilin-azo-meta-tolylene-diamin. Shade. — Yellow brown. Offered by 1 out of 12 sources. 41. Trade name. — Chrysoidin R. Scientific name. — Hydrochlorid of toluene-azo-meta-tolylene-di- amin. Discovered. — 1876. Shade. — Orange Brown. (Note. — It has not been possible accurately to differentiate in the literature as to whether Nos. 17, 18, or 41, or all three, were referred to.) FAVORABLE. 1. V/eyl (p. 115): " Nonpoisonous * * * Chrysoidin * * *. " 2. Permitted by the law of Italy. UNFAVORABLE. 1. Prohibited by Confectioners' List. 2. Weyl (p. 126): Experiment on dog, giving him 113 milligrams per kilogram body weight, or 79 grains per 100 pounds body weight, produced a slight albuminuria. In a second experiment a dog receiving "1 grain" ("grain"' is evidently a misprint for "gram" and will be so treated), that is, 105 milligrams per kilogram body weight, or 74 grains per 100 pounds body weight, daily for one month, did not produce albuminuria, but caused a lossof body weight of about 12.5 percent. A third dog receiving about 2 milligrams per kilogram body weight, or about 1.5 grains per 100 pounds body weight subcutaneously, suffered a loss of 20 per cent of its body weight in 23 days; dose repeated 24 days after and animal kept under observation seven days longer when animal was normal. 3. Weyl in summarizing experiments on these three dogs concludes as follows: "Chrysoidin produces, according to my investigations, a slight albuminuria, and notable reduction in body weight, but further disturbance has aol been noted." {p. 127.) 4. Fraexkel (/). 577): "The abovo-iiKiitioned Chrysoidin * * * produ slight albuminuria, and a notalAc decrease of body weight, and producet factory eczema." 5. Lewin (Lekrbuch der Toxikolo; p. IS1)\ " Produces eczema/' and cites Deutsche Med. Wnchatschr., 1891, p. 45. G. T. 28. Trade name. — Archil Substitute Y. Sen ittific naiitt . — Sodium salt of |>ara-nit lolxMixeiie-a/o-alplia-na j>h- bhylamin-para-sulphonic add. Discovi red and paU nted.—l&7&. Shwl< . — Red. Not offered. 88 COAL-TAR COLORS USED IX FOOD PRODUCTS. FAVORABLE. 1. Permitted by Confectioners' List. 2. \Veyl(p. 115): "Nonpoisonous * * * Archil Substitute * * *." Describing experiments on three dog3, as follows: A. 430 milligrams per kilogram body weight, or 301 grains per 100 pounds body weight, on each of two successive days, and double the dose on the fourth day, producing no vomiting, but a tendency to vomit, a slight albuminuria and colored urine. B. 182 mil- ligrams per kilogram body weight, or L27 grains per 100 pounds body weight, administered daily for one month; results similar to foregoing, but no colored urine. C. 105 milligrams per kilogram body weight , or 116 grains per 100 pounds body weight administered subcutaneously, produced only slight albuminuria, and no reduction in body weight, (p. 125.) G. T. 43. Trade names. — Orange GT; Orange RN; Orange O; Orange N. Scientific name. — Sodium salt of toluene-azo-beta-naphthol-sulpho- nic acid. Discovered. — 1879. Shade. — Orange. Not offered. FAVORABLE. 1. Permitted by Confectioners' List. UNFAVORABLE. 1. Excluded by Austrian law. 2. Excluded by Swiss laws. G. T. 55. Trade names. — Ponceau R; Ponceau 2 R; Ponceau G and GR; Xylidin Red; Xylidin Scarlet. Names under which it was offered on the United States market as a food color in 1907. — Scarlet; Orange R. Scientific name. — Sodium salt of xylene-azo-bcta-naphthol-disul- phonic acid. Shade. — Scarlet. Offered by 2 out of 12 sources. FAVORABLE. 1. Permitted by Confectioner-' List. 2. Weyl (p. 31): "According to Cazeneuve ami Lepine'e experiments the follow- ing are not poiaonous to human beings and doge * * * Ponceau R * * *." 'A. "Ponceau K (Ponceau 2 R, Xylidin Red, Xylidin Ponceau), not poisonous to dogs neither by administration by stomach n<>r injection into blood." (p. 115.) 4. "Other Azo colors * * * Xylidin Red * * * are entirely nonpoisonous." {p. lis.) :>. Fbainxbl(p.575): "Thatthe monazo coloring matters examined by Caseneuve mid I /pi n<\ as already above stated, were nonpoisonous, can be easily explained by the constitution <>f these substances. These two investigators examined * * * Ponceau It * * *." COMPILED DATA UNDER GREEN TABLE NUMBERS. 89 6. Lieber (p. 140): A guinea pig received 310 milligrams per kilogram body weight or 217 grains per 100 pounds body weight, once a day six times ever}- other day; the appetite appeared to remain good, and no disturbances were noted. 7. Permitted by Austrian law. 8. Permitted by Swiss laws. 9. Cazexeuve and Lepixe (Bull, de Vacad. de mid., 1886, p. 643): Tolerated by man well or sick. 10. Chlopin {p. 150) classes it as nonpoisonous on his own experiments. His experi- mental data are as follows: Experimental data by Chlopin. [1 gram=141 mg= 99 grains.] Date. Dose. Weight. 24-hours' urine. General condition of animal and urine. 1902. May 3 Grams. 2 Kilos. 7.1 cc. 370 351 420 290 359 360 330 Dog and urine normal and no albumen. Urine rose-colored; no albumen. 5 2 Color normal; no albumen. 6 Rose-colored urine; no albumen. 7 Normal color; no albumen. g Rose-colored; no albumen. 9 Color and composition normal. Total. 6 UNFAVORABLE. Prohibited by the ordinance of the police commissioner in France. (See Lieber, p. 30.) Meyer (J. Amer. Chem. Soc. 1907, v. 29, pp. 900-901): The dog experimented on showed signs of paralysis on the morning of the seventh day at 8 o'clock, and died at 10.40 a. m., after having received a total of 32 grams of color, of which 1G had been given on the last day. The initial dose was 70 grains per 100 pounds body weight; the total weight of color was 5,818 milligrams per kilogram body weight, or 4,073 grains per 100 pounds; the average dail] was therefore 582 grains per 100 pounds, or 831 milligrams per kilogram body weight. DOUBTFUL. 1. Winogradow (Zts. Xahr. Genussm. inhibits digestion, 1903, '. 5, p. 689) says it almost completely G. T. 65. Trade names.— Fael Red Bj Bordeaux I>: Bordeaux BL; Bordeaux K extra. Names under which it woe offered on (lit United States market as a d color in 1907* Bordeaux B; Claret lvc«i. Scientific name.-— Alphanaphthj Uanin-azo-betanaphthol-disulpho- oic acid. Discovered. 1878. Shade. Red. Offered by 2 out <»f L2 bout 90 COAL-TAR COLORS VSED IN FOOD PRODUCTS. FAVORABLE. 1. Permitted by Confectioners' List. 2. Weyl (p. 115): " Bordeaux Red (Fast Red B) not poisonous to human beings." 3. "Other Azo colors * * * Fast Red B * * * are entirely nonpoisonous." (p. 148.) 4. Fraexkel (p. 575): "That the monoazo coloring matters examined by Cazeneuve and Lepine, as already above stated, are nonpoisonous, can be easily explained by the constitution of these substances. These two investigators examined * * * Bordeaux B." 5. Arloixg and Cazeneuve (Archives de physiologie, 1887, pp. 856-393): As a result of this work, which is divided into three parts — (1) Stating the effect of direct introduction of the color into the circulation; (2) intravenous injections; (3) comparing the effects of injections of color and of salt; and (4) feeding by the mouth — these investigators conclude that these coloring matters are toxic only in extremely large doses; that when given to dogs with their food that no incon- venience of any kind results; this is based upon experiments on three dogs, covering 145 days, where each dog received per kilogram of initial body weight in the first case 20,307 milligrams, or 14,213 grains per 100 pounds initial body weight; in the second case, 29,590 milligrams, or 20,713 grains per 100 pounds initial body weight; and in the third case, 28,154 milligrams, or 19,758 grains per 100 pounds initial body weight. Per day this means 98 grains per 100 pounds initial body weight in the first case; in the second case, 143 grains per 100 pounds initial body weight; and in the third case, 137 grains per 100 pounds initial body weight. 6. Cazeneuve (Arch. gen. de med., 1886, p. 753) says it may be taken without effect by man or animals, sick or well, in large doses. 7. Cazeneuve and Lepine (Bull, de Vacad. de med., 1886, p. 643): Tolerated by man well or sick. G. T. 70. Trade name. — Azarin S. Scientific name. — Ammonium bisulphite compound of dichloro- phenol-azo-beta-naphthol. Shade. — Red. Not offered. FAVORABLE. 1. Permitted by Confectioners' List. 2. Weyl (p. 115): "Nonpoisonous * * * Azarin S * * *. " xperiments on five dogs; three fed by the mouth; two treated hypodermically, of which latter one died. The first dog received 1,307 milligrams per kilogram body weighl in 25 days; that is, 54.7 kilograms per day on the average, or a total of 957 grains per 100 pounds of body weigh! ; that is, 38 grains per 100 pounds body weight per day. The Becond dog received a total of 1,942 milligrams ]>or kilogram body weighl in 20 days, orJ)7 milligrams per kilogram body weight per day, which amounts to a total of 1,359 grains per LOO pounds body weight, or 68 grains per I')') pounds body weighl per day. In both Cases a distinct amount of albumen was present in the urine, and the urine 6 vol ved sulphurous acid on treat- ment with hydrochloric acid. The third dog received hypodermically three do,- in eight days, each dose being 213 milligrams per kilogram bodj weight; that i-. I 19 '/mil!- per l<»" pounds body Weight) no had effects. (/'• !■■■■ I 3. "Administered by the stomach A/.arin S is harmless." (p. 134.) 4. " Other Azo colors, * * * for instance Azarin 8, are entirely nonpoisonous. M (p. 148.) .', PraBNKBL \/.;irm S administered hv I ho stomach i- entirely harmless. " COMPILED DATA UNDER GREEN TABLE NUMBERS. 91 UNFAVORABLE. 1. Weyl (p. 134): Dog; weight not given; received 5 cc of Azarin S paste by injec- tion into the abdominal cavity, and survived three days. "The cause of death was considered to be peritonitis without effusion. The result of this post- mortem is of much interest. The red spots consisted, as was determined by chemical analysis, of the azo color which is the basis of the Azarin S. Conse- quently in the peritoneal cavity the same splitting up of the Azarin S had occurred which takes place when it is attached to textiles. " 2. Fraenkel (p. 578): To the same effect. G. T. 78. Trade name. — Erika B. Scientific name. — Sodium salt of methyl-benzen} T l-amido-thio- xylenol-azo-alpha-naphthol-disulphonic acid . Discovered and patented. — 1889. Shade. — Rose Pink. Xot offered. Nothing. FAVORABLE. UNFAVORABLE. 1. Chlopin (p. 153) : Based on his own experiments considers it as not harmless. The experimental data are as follows : Experimental data by Chlopin. [1 gram=125 mg=87.5 grains.] Date. Dose. Weight. 24 hours' urine. General condition of animal ami urine. 1901. 10 Grams. 2 Kilos. 8.0 cc. 392 400 120 300 390 293 J60 no 102 880 Dog normal; mine add; no albumen. brine of rose shade; insignificant traces of albumen; acid. 11 Color same; no albumen. 12 2 Do. u Do. 11 Urine vellow, greenish shade; traces of albumen. 15 17 2 Do. Urine wine vellow, orange, acid; DO albumen. 19 Dog quite well; urine orange: no albumen. Total . 6 G. T. 84, Tradt names. — Resorcin Yellow; Tropeolin Oj Trop®olin K: Chrysoin; Chiyseolin; Yellow T; Gold Yellow; Acme felloe , Names under which it was offered on the United States market as q color in i'«>7. Chrysoiii REZj Resorcio 0275. Scu ntific nam . Sodium Ball of para-eulphobenzene-azo-resorcinol. Discovered. L875. Shadi . — Reddish vellow. Offered by 2 out of 12 sources. 92 COAL-TAR COLORS USED IN FOOD PRODUCTS. FAVORABLE. 1. Chlopin {-pp. 131-2) examined this color physiologically, and has classified it as harmless. The experimental data are as follows: Experimental data by Chlopin. No. l (p. M7). [1 gram=43 mg=30 grains.] Date. Dose. Weight. 24 hours' urine. General condition of animal and urine. 1901. Apr. 9 10 Grams. Kilos. 23.00 cc. 500 Dog quite well; urine normal color, acid; no albumen. Do. 11.21 2 0.70 12.00 11 Urine acid; no albumen. 12 No symptoms of poisoning; eats. Urine dark brown; no albumen. 13 460 520 580 14 Do. 17 Urine normal color; no albumen; dog is well. Total . . 13.21 2 0.70 No. 2. [1 gram=156 mg=109 grains.] 1901. Oct. 9 10 11 12 13 14 15 16 17 18 19 20 21 22-26 Total... 6.4 30 350 2 2 300 335 290 290 375 300 2 2 2 320 360 360 7.0 10 Dog well; urine yellow; no albumen. Do. Urine brown, acid; no albumen. Urine light brown, acid; no albumen. Do. Do. Do. Do. Do. Do. Urine light brown, acid, no albumen; diarrhea. Brown yellow, acid, no albumen no diarrhea. Urine normal; dog is well. Do. 1 Internally. 2 Subcutaneously. 2. Permitted by the law of Italy. UNFAVORABLE. 1. Forbidden by the Confectioners' List. DOUBTFUL. 1. Winogradow (Zts. Nahr. Genussm., 1903, v. 6, p. 589) says it noticeably retards digestive action; not indifferent. G. T. 85. Tradenames. — Orangel; Alphanaphthol Orange; Naphl bo] Orange; Tropseolin OOO; Orange B. Names under which it was offered on the United States market as a food color in 1007. — Orange KZ; Orange 027. Scientific name. — Sodium salt of para-sulphobenzene-azo-alpha- oaphthol. Discovered, — 1876. Shade. — Orange. Offered by 2 out of 12 sources. COMPILED DATA UNDER GREEN TABLE NUMBERS. 93 FAVORABLE. 1. Permitted by Confectioners' List. 2. Weyl (p. 31): "According to Cazeneuve and Lepine's experiments, the following are not poisonous to human beings and dogs: * * * Orange * * *." 3. "Orange I (Alpha-naphthol Orange, Tropaeolin OOO) not poisonous to dogs neither by administration by stomach, nor by injection into blood." (On authority of Cazeneuve and Lepine.) (p. 115.) 4. Weyl {pp. 123, 148) refers to this as not poisonous. 5. Permitted by the law of Italy. 6. Permitted by the law of Austria. 7. Cazeneuve (Arch. gen. de med., 1886, Vol. I, p. 753) says it may be taken without effect by man or animals, sick or well, in large doses. 8. Cazeneuve and Lepine (Bull, de I'acad. de med., 1886, p. 643): Tolerated by man, well or sick. G. T. 86. Tradenames. — Orange II; Betanaphthol Orange; Tropaeolin 000 No. 2; Mandarin G extra; Chrysaurein; Gold Orange; Orange extra; Atlas Orange; Orange A. Names under which it was offered on the United States market as a food color in 1907. — Xaphthol Yellow SLOZ; Orange II; Orange; Orange Y; Mandarin G extra; Orange A 1201; Orange A extra. Scientific name. — Sodium salt of para-sulphobenzene-azo-beta- naphthol. Discovered. — 1876. Shade. — Orange. Offered by 8 out of 12 sources. FAVORABLE. 1. Permitted by law in Italy. 2. Frentzel (Zts. Nahr. Genussm.,1901, < . 4, p. 974) says that according to his experi- ments this color, in the small amounts in which it is used in food products and which can enter the human system in the course of 24 hours, canhardh with frequent administration, cause a harmful affect 3. Frentzel (Zts. Nalir.Genussm., 1901, v. 4. pp. 968-974)'- Experimented on rabbits, giving dye with food, a total of 21 grams in 19 days, ■ total of S.748 milligram* per kilogram body weight, or (i, 133 grains per 100 pounds body weight, in of 1 gram each, daily for the first 15 days, or about 417 milligrams per kilogram of body weight, or 292 grains per 100 pounds body weight pel doae far these L5 doses; the color could only l>e detected in the urine, and th< • came softer. A dog was given i,o20 milligram! per kilogram body weight, Of 714 grains per loo pounds body weight, and showed distinct kidney irritation, great thirst, and diarrhea; recovery required about one week; and thereafter the same animal was fad by the mouth one-iw ant ieth ol the above d0S8 cadi day for Boccessh e days without any untoward effect , A second dog recede ed per kilo- gram of body weight 172 milligrams, or 121 graini per LOO pounds body weight, and it, like the lirst dog, in tin- first experiment, showed kidney irritation, diarrhea, and great thirst. < >n humans K>0 milligrams, or U grains, colored the urine within b"> minutes, and 1 his color remained far 24 hours; there 1 Sign "f Vomiting or diarrhea; the hitter taMe q| the COloi 1 able, 94 COAL-TAR COLORS USED IX FOOD PRODUCTS. UNFAVORABLE. 1. Forbidden by Confectioners' List. 2. Forbidden by Swiss Analytical chemists. 3. Wbtl (p. US): "Poisonous Orange II. * * *." 4. " Betanaphthol Orange is, therefore, according to Experiment I, poisonous in small doses when administered by the stomach, and suffices to kill an ordinarily large strong dog." (p. 123) 5. A. Weyl's own experiments on two dogs, the initial dose in one case being 476 milligrams per kilogram body weight, or 333 grains per 100 pounds body weight; the animal receiving in the course of 20 days, in four doses, 1,333 milligrams per kilogram body weight, or 933 grains per 100 pounds body weight; or 335 milligrams per dose per kilogram body weight; that is, 225 grains per dose per 100 pounds body weight. The animal died, and suffered diarrhea and albumi- nuria, and its urine was colored red throughout the entire period. B. The second dog received hypodermically per kilogram body weight, 116 milligrams, or 61 grains per 100 pounds body weight; its urine was colored; albuminuria, diarrhea, loss of hair, abscesses, and loss of weight occurred. It required 36 days to recover from four doses administered during one week. C. A rabbit received 1,333 milligrams per kilogram body weight, or 933 grains per 100 pounds body weight, and died within 12 hours, (p. 122) 6. "Of the 23 Azo colors subjected to examination only two * * * Orange II produce(s) such effects when administered by the stomach that we can con- sider it poisonous. With dogs the lethal dose is less than 1 gram per kilo of the body weight of Orange II * * *." (p. 147.) 7. "Further Orange II, which is poisonous * * V 8. "The poisonous qualities of Orange II." 9. "Further, in spite of the presence of the sulpho groups, colors may be poisonous, as is shown with Orange II." (p. 148.) 10. Chlopin (Zts. Nahr. Genussm., 1902, v. 5, p. 241): A. A dog received 349 milli- grams per kilogram body weight, or 244 grains per 100 pounds body weight for the first day of experiment, which dose was repeated on the third and fourth days; nothing untoward is noted for the first three days in the condition of the dog; the urine was dark red but free from albumen; on the fifth and sixth days two-thirds of the above amount was given, and on the seventh and eighth days the original dose was given. On the fourth day the animal was frisky and had a good appetite but was vomiting; on the fifth day vomiting stopped, but diarrhea ensued, which diarrhea continued for one week; the weight remained practically constant; the urine was colored throughout from dark red to orange red and dark brown and became normal the fourth day after the last ad mi nisi ra- tion. B. Humans: Chlopin took 200 milligrams, or 3^ grains, in a gelatin capsule at 3 p. m.; at 4.30 p. m. the urine was colored a strong red orange; at 6 p. in. a dryness of throat and bad taste in mouth appeared; at 6.30 p. m. felt very badly; vertigo and unable to remain seated and continue writing; blood rushed to head; tin- general condition very poor; somewhat improved by moving about in open air; 7.30 felt so poorly took (dauber's -alt as an antidote; II |». m. ill condition still continuing; urine normal yellow; midnight recovered. Chlopin states that he would not repeal this experiment on himself, or on any other human, and he concludes therefore that this color must he regarded as harmful. 11. CnLOPiN (p. 1.33) (dasses it as "harmful." The experimental data are as follows: COMPILED DATA UNDER GREEX TABLE NUMBERS. 95 Experimental data by Chlopin. [1 gram=116mg=81 grains.] Date. Weight. 24-hours urine. General condition of animal and urine. 1901. May 27 28 29 30 31 June 1 2 3 4 5 Total.. Grams. Kilos. 8.6 8.2 8.59 Dog is quite well; urine normal. Urine brown red, acid; no albumen. Do. Diarrhea and vomiting at night, no albumen; urine orange red, acid, no albumen. Diarrhea continues; no albumen; no vomiting. Diarrhea continues; dog lively, eats with relish; no albumen. Urine brown orange, acid, no albumen; diarrhea continues. Urine clear; no albumen; diarrhea continues. Urine dark brown; no albumen; diarrhea decreases. Diarrhea still less: no albumen. Diarrhea very slight; urine Slightly orange, no albumen. Urine normal in color and composition; diarrhea stopped. 12. Buss lists it as poisonous. DOUBTFUL. 1. Winogradow (Zts. Xuhr. Genussm., 1903, v. 6, p. 5S9) says it almost completely inhibits digestion. G. T. 87. Trade names. — Orange III; Ilolianthin; Tropaeolin D; Methyl Orange; Dimetlivl.-milin Orange. Scientific name. — Sodium salt of para-sulphobenzene-azo-diniethy- lanilin. Discovered. — 1876. Shade. — Orange Yellow. Not offered. FAVORABLE. 1. Permitted by the law of Italy. IN FAVORABLE. 1. Chlopin (pp. 145, 146) on his own experiments classes it as poisonous. The experimental data are as follows: Experimental data by Chlopin. No. 1. [1 gram = l(V{ mg = 114 grains.] ' "Trim'"' General condition of animal :m. Sept. 1901. M M 1 2 :< ■» 5 7 Grams. Kilns. <> 6b U 310 805 •J a Total .. 10 Before experimenl dog quite normal; color urine usual, add, no albumen. Do. Color of urine dark yellow; no albumen. Urine brown; reddish with sulpburio add; no albumen. Do. Do. Urine dark brown; no albumen; vomit dog does not eat, but drinks with Complete ataxia, which became general on the 9th; dog can do! move tn straignt line, and wa on tl lis; after Calling arises with difficulty and with widely spread feet, continuously lifting r then the other; extremii on 11th days I cultv; placed on his feet, maintained equilibrium with dif- ficulty: involuntarily H 1 moves it about in the air; animal can stlU swallow; i I to light; d id the L2th the hind legs completelj | on the ltth vomiting ::■ of the fro: • quietly without moving; La kill . shows no change in Lri- i i in the lumbar repion of th<> ■ .1 column (?) borderinj nterior and Lateral columns," causing death by paralysis of the heart. 96 COAL-TAR COLORS USED IN FOOD PRODUCTS. Experimental data by Chlopin — Continued. No. 2. [1 gram=149 mg= 104.3 grains.] Date. Dose. Weight. 24 hours' urine. General condition of animal and urine. 1901. Dec. 3 Grams. 3 Kilos. 6.7 cc. A few hours after giving color dog in tremors; does not eat. Urine black; acid; no albumen; paralysis of the extremities. Dog lies in cage in full paralysis; died "before dinner; no albu- men in urine; cause of death, paralysis of heart. 4 Little. Little. 2. Meyer (J. Amer. Chem. Soc, 1907, v. 29, p. 900): Dog receiving 113 milligrams per kilogram body weight; that is, 79 grains per 100 pounds body weight. Result, diarrhea, which continued throughout 17 days, although only 3 doses of the same size were given in 6 days; thereafter the dose was increased geometrically; the diarrhea continued; the coloring matter was discharged in the urine and the feces; there was no abnormal condition revealed by the autopsy. DOUBTFUL. 1. Winogradow (Zts. Nahr. Genussm., 1903, v. 6, p. 589) says it noticeably retards digestive action; is not indifferent. G. T. 88. Trade names. — Diphenylamin Orange; Orange IV; Tropseolin 00; Orange M; Fast Yellow; Orange G S; New Yellow; Orange N; Acid Yellow D. Scientific name. — Sodium salt of para-sulphobenzene-azo-diphenyl- amin. Discovered. — 1876. Shade. — Orange Yellow. Not offered. FAVORABLE. 1. Weyl (p. 115): "Nonpoisonous * * * Diphenylamin Orange * * * 2 * * * Diphenylamin Orange is * * * nonpoisonous." (p. 132.) 3. "For instance, the poisonous Metanil Yellow corresponds to the nonpoisonous Diphenylamin Orange." (p. 148.) 4. The experiments on which Weyl based the conclusions above may be summarized as follows: A. A dog received 183 milligrams per kilogram body weight, or 128 grains per 100 pounds body weight. The urine was rendered black, and contained traces of phenol and abundant albumen; this condition lasted for 72 hours, at the end of which 110 milligrams per kilogram body weight, or 77 grains per 100 pounds body weight, were administered; urine became colorless and albumen diminished. Five days afterwards a total of 10 grams, or 366 milligrams per kilogram body weight, or 256 grain- per 100 pounds body weight, were administered; the urine became abundant, was strongly black, alkaline, contained albumen. The loss qJ freight was 1 kilogram, or about 3$ percent. B. A dog received 308 milligrams per kilogram, or 216 grains per LOO pounds body weight, as the initial dose; albuminuria did not result until after repeated dosing same ai initial dose, and administered for 2 weeks. 5. Weyl'.- conclusions arc as follows: "According to (he above investigations, Diphenylamin Orange causes albuminuria, but further disturbances did not appeal during the Beveral weeks' observations on the animals used," COMPILED DATA UNDER GREEN TABLE NUMBERS. 97 6. Chlopin (p. 148) examined this color and classes it as nonpoisonous. The ex- perimental data are as follows : Experimental data by Chlopin. [1 gram=145 mg=100 grains.] Date. Dose. WdtfJ "Jgf General condition of animal and urine. 1901. Oct. 27 28 Grams. 3 Kilos. 7 cc. 300 Dog quite normal; urine color normal; no albumen. Urine dark brown; acid; no albumen. Do. Do. Do. Urine yellow, with sulphuric acid red; no albumen. Orange with orange sheen; no albumen. Do. Urine brown vellow: acid: no albumen. 29 3 3 3 3 3 3 30 31 Nov. 1 2 340 3 G.5 4 275 5 3 300 Do. 6 6.7 Do. Total. 24 7. Fraenkel (pp. 577, 578) to the same effect as Weyl, as above. 8. Permitted by the law of Italy. UNFAVORABLE. 1. Prohibited by Confectioners' List. G. T. 89. Trade names. — Brilliant Yellow S; Yellow WR; Curcumin. Names under which it was offered on the United States market as a food color in 1907.— Brilliant Yell. >w S. Scientific name. — Sodium salt of para-sulphobenzenc-a/.o-diphenyl- amin-suphonic acid. Shade. — Yellow. Offered by 1 out of 12 source-. 1 AVORABLE. 1. Lieber (p. 136): A very young rabbit received ,-ix doses on alternate days, doae amounting to 320 milligrams per kilogram of body weight, or 22 1 grains per 100 pounds of body weight. No untoward symptoms arc recorded; the body vreight increased almost 10 percent in 11 da Nothing. i \ ! A\ ORABLE. G. T. 92 Tradi minus. Ax<> Acid Yellow; A/«>ila\in: Azo Yellow; Indian Yellow. Scientific name. Mixture of nitrated diphenylamin yellow with nit n>-diplieii\ lamin. Piston in/. I^so. sIkk/i . Yellow. Not offered. • 1° Hull. 117—12 7 98 COAL-TAR COLORS USED IN FOOD PRODUCTS. FAVORABLE. 1. Chlopin (p. 128) examined this color and classifies it as nonpoisonous. The ex- perimental data are as follows: Experimental data by Chlopin. No. l. [1 gram=125mg=87.5grains.] Date. Dose. Weight. urine^ I General condition of animal and urine. 1901. Apr. 6 8 Grams. 2 Kilos. 8 cc. 370 Little. Dog normal; urine acid; no albumen. Urine reddish orange; acid; no albumen. Do. 10 3 11 600 430 Urine red; no albumen. 12 Urine almost normal color; acid; no albumen. Total . 5 Conclusion: Nonpoisonous. No. 2 (p. 129). [1 gram= 143 mg= 100 grains.] Date. Dose. Weight. 24 hours' urine. General condition of animal and urine. 1901. Apr. 12 14 Grams. 2 3 Kilos. 7 cc. 430 610 600 400 Dog normal; urine acid; no albumen. Do. 15 Urine dark vellow; acid; no albumen. 16 Urine almost normal color; acid; no albumen. Total.. 5 Conclusion: Showed no harmful effects. 2. Permitted by law in Italy. UNFAVORABLE. Nothing. DOUBTFUL. 1. Winogradow (Zts. Nahr. Genussm., 1903, v. 6, p. 589) says it almost completely inhibits digestion; noticeably retar.l- digestive action; is not indifferent. G. T. 93. Trade name. — Azo-fuchsin G. Scientific name. — Sodium salt of para-sulphobenzene-azo-dioxy- naphthalene-sulphonic acid. Discovered. — 1880. Slwde. — Reddish brown. Not offered. FAVORAHLK. l Chlopin (pp 1 >>. i 7) examined this color and classifies it as nonpoisonous. The experimental data an- at foil* COMPILED DATA UNDER GREEN TABLE NUMBERS. Experimental data by Chlopin. No. 1. [1 gram=117 mg= 82 grains.] 99 Date. Dose. Weight. 24 hours' urine. General condition of animal and urine. 1901 May 3 Grams. 2 Kilos. 8.5 cc. 300 363 390 380 285 420 292 Dog well and normal; urine acid; no albumen. Urine chocolate brown; acid; no albumen. 5 2 Do. 6 Urine dark brown, chocolate, in thin layer greenish sheen; acid; no albumen. Urine slightly greenish; acid; no albumen. 7 2 8 Dark chocolate brown; acid; no albumen. 9 8.7 Urine normal; dog is well. Total. 6 No. 2. [1 gram=125mg= 87.5 grains.] 1901. Oct. 9 8.0 390 310 10 2 2 11 12 380 325 13 2 14 15 2 2 396 16 17 397 325 380 246 420 18 2 19 20 21 8.4 Total. 12 Dog quite normal; urine acid; no albumen. Do. Urine slightly more yellow; no albumen. Urine dark brown; acid; no albumen. Do. Do. Do. Do. Do. Do. Urine almost black; acid; no albumen. Urine almost black; acid; no albumen; urine of lighter color. Do. Nothing. UNFAVORABLE. DOUBTFUL. 1. Winogradow (Zts. Nahr. Genussm., 1903, v. 6, p. 589) says it almost completely inhibits digestion. G. T. 94. Trade names. — Tartrazin; Hydrazin Yellow. Names under which it was offered on the United States market as a food color in 1907. — Flavazein Red Shade Z; Tartrazin; Acid Yel- low AT. Scientific name. — Sodium salt of benzcne-azo-pyrazalone-caiboxv- disulphonic acid. Discovered. — 1884. shade. — Yellow. Offered l>y 6 <>ut of L2 lources. DOUBTFUL. 1. Lieber (p. 134): Dog, 2 months old, received, pet kilo, body weight, :'.: milli- gram-, <>r 26 gnim per i<><> pounda body weight, -ix rimon on a lt ernate day* The animal sul'lVrod (TOO) mild diarrhea at the .-tart, which Continued with greater or less activity than at tin- -tart throughout thot<-t. 100 COAL-TAR COLORS USED IX FOOD PRODUCTS. 2. Meyer {J. Amer. Chem. Soc, 1907, v. 29, p. 897): Dog received 100 milligrams per kilogram body weight, or 70 grains per 100 pounds, increased in geometric proportion for 6 consecutive days, at the end of which time diarrhea set in. On that day 2,000 milligrams per kilogram body weight, or 1,400 grains per 100 pounds body weight, were administered; this was about two-thirds as much as the animal had received in all the 5 days preceding; the feces were colored after the first administration, and the urine was also colored; albuminuria doubtful. 3. Fraenkel (pp. 210 and 216): Tartrazin, according to the Green Tables, is a derivative of isopyrazolon; "* * * only those substances which are deri- vatives of pyrazolon are antipyretics, the isopyrazolon derivatives are, how- ever, poisonous." Tartrazin is also closely related to the antipyrin class of compounds, which class is known to possess an irritant action and also a de- pressing action on the circulation. Tartrazin also contains benzol groups, which are said to increase the physiological activity of pyrazolon derivatives. G. T. 95. Trade names. — Metanil Yellow; Orange MN; Tropaeolin G. Names under which it was offered on the United States market as a food color in 1907. — Victoria Yellow cone. Z; Yellow MXX cone. Scientific name. — Sodium salt of meta-sulphobenzene-azo diphenyl- amin. Discovered. — 1 8 7 ( . I . Shade. — Orange. Offered by 2 out of 12 sources. FAVORABLE. 1. Frentzel (Zts. Nahr. Genussm., 1901, v. 4, p. 974): A. A rabbit received 379 milligrams per kilogram body weight, or 265 grains per 100 pounds body weight; no color administered for 2 days, and then administered at intermittent periods, so as to receive 6 doses in 19 days. Total weight administered per kilogram body weight 2,085 milligrams, or 1,400 grains per 100 pounds, whieli is equivalent to 77 grains per day per 100 pounds; the color could always be detected in the urine, and the feces became softer. B. A dog received 581 milligrams per kilogram body weight at one dose; that is, 407 grains per 100 pounds body weigh! . There was no vomiting and no diarrhea; the dye persisted in the urine for 72 hours. C. The same dog, after recovering from the foregoing, received daily one-tenth the above dose for 9 days; the dyestufi was found in the urine and the feces and movements were normal; dissection showed nothing abnormal. I». A human swallowed loo milligrams, or L.g grains; no untoward symptomi are recorded; tin- urine remained colored for 24 hours. From the foregoing Pren'tzel Concludes that this coloring matter is absolutely harmless. 2. Chlopin [Zti. Nahr. Qenuttm., A. A dog received 305 milli- • , per kilogram body weight; thai is, L04 grains per LOO pounds body weight; no color administered for four days; the same dose was given on alternate days mx times, and the dose was increased 60 per cent, and that dose administered twice, and the original dose was given on the next alternate day; altogether the animal received per kilogram body weight, 3,355 mUligrams, or 1,144 grains per LOO pounds body weight, in i period of 22 days, or an average o! 153 milli- gram! per kilogram body weight per day, or 52 grains per day per LOO pounds body weight. Ai tie- end of the twelfth day albuminuria set In, and it required throe m i k iftei < et ing the administration of the color for the albuminuria COMPILED DATA UNDER GREEN TABLE NUMBERS. 101 2. Chlopin (Zts. Nahr. Genussm., 1902, v. 5, p. 241 )— Continued. to disappear; the urine was colored throughout the period of dosing, and it was not until 10 days after the last color had been administered that the urine regained normal color; otherwise the animal was well throughout. B. Hu- mans: Two hundred milligrams, or 3-T2 grains taken at 3 p. m., October 30, 1901, colored the urine so highly yellow that it could be dyed with. Apart from the bitter taste of the product no untoward symptoms are recorded. C. Chlopin's conclusion {Zts. Nahr. Genussm., 1902, v. 5, p. 244)'- "Metanil Yellow is not poisonous to dogs in doses of from 2 to 3 grams per day, nor to humans in doses of 0.2 grams per day, and may therefore perhaps be regarded as non- poisonous from a practical point of view." UNFAVORABLE. 1. Weyl (p. 115): "Poisonous * * * Metanil Yellow." 2. Says the product smelled strongly of diphenylamin. (/;. ISO.) 3. A. "Metanil Yellow must be considered poisonous when administered by the stomach from the indications of Experiments 1 and 2. The lethal dose, which is determined by Experiment 2, is 0.53 grams per kilo body weight." This lethal dose is 371 grains per 100 pounds body weight. B. A dog received 862 milligrams per kilogram body weight, or 603 grains per 100 pounds body weight. This caused vomiting; the same dose was repeated 24 hours afterward.-*, the animal again vomiting. The animal died within 96 hours from the first admin- istration. C. A dog received 89 milligrams per kilogram body weight, or 62 grains per 100 pounds body weight; the urine became colored about 96 hours after administration; 5 days after the first administration the animal was given 10 times the original dose; vomiting set in within one hour; in 24 hours the urine was deeper colored; a week later after the dose last preceding, one-half of that dose was given, and the animal died within 24 hours; the animal had lost during this period approximately one-quarter its original weight, (p. 182.) 4. "* * * Metanil Yellow * * * produce[s] such effects when administered by the stomach that we can consider them [it] poisonous." (p. 147.) 5. "The poisonous qualities of * * * Metanil Yellow; the poisonous Metanil Yellow." (p. 148.) 6. Fraexkel (p. 578): "A dog weighing 11 kilograms was killed by 20 grams of this coloring matter within four days, whereas the isomeric Diphenylamin Orange is nonpoisonous, and it must first be considered if the poisonous nature of this substance can be explained by the easy liberation of diphenylamin from it, since this coloring matter, in and of itself, ha- a strong odor of diphenyla- min." 7. Chlopin (p. 141)'- Based on his own experiments considers it "Not quite harm- fas." The experimental data arc as follows: imental data fry chlopin. [1 Kram 153 XQg In? grains.] D ito. Weight. M boars' urine. (.fin r.il < cm id it ion <>f animal and urine. 1901. M.r. Lfl 1 | cc. ISO 100 370 ( rine unusual color; acid; no albumen. Do 20 21 I'rii ■ 1; DO albumen. Do Do 24 B0 25 M 2 27 102 COAL-TAR COLORS USED IN FOOD PRODUCTS. Experimental data by Chlopin — Continued. Date. Dose. Weight. 24 hours' urine. General condition of animal and urine. 1901. May 28 29 Grams. 2 Kilos. cc. 330 350 350 355 400 340 330 400 305 390 330 335 Do. 30 2 Do. 31 Do. 2 2 Do. * 3 Do. 4 3 Do. 5 Do. 6 3 Do. 7 Do. 8 2 Do. 9-29 6.5 Color of urine normal June 18, on 29th albumen disappears. Dog is well. Total. . 22 8. Chlopin (p. 141)'- On authority of others not stated, classes this color as harmful or poisonous. 9. Prohibited by the laws of Italy. 10. Prohibited by Confectioners' List. 11. Prohibited by the Resolutions of the Society of Swiss Analytical Chemists, Sep- tember, 1891. 12. Buss lists it as poisonous. DOUBTFUL. 1. Winogradow {Zts. Nahr. Genussm., 1903, v. 6, p. 589) says that it noticeably retards digestion. G. T. 97. Trade names. — Orange T; Mandarin G R; Orange R; Kermesin Orange. Name under which it was offered on the United States market as a food color in 1907. — Orange 2 R. Scientific name. — Sodium salt of sulpho-ortho-toluene-azo-beta- naphthol. Shade. — Orange. Offered by 1 out of 12 sources. Nothing. FAVORABLE. UNFAVORABLE. 1. Chlopin {p. 125): Based on his own experiments considers this color as "harm- ful." The experimental data are as follows: No. l. D:ii.-. 1 >■< '■. Weight. 24 hours' urine. General condition of animal and vim. 1901. Apr. 4 Grams. Kilos. cc. 310 280 420 570 :i7o an 330 Nothing abnormal. Urine reddish brown; traces of albumen; acid. 6 3 Do. 7 I'rine dark brown; acid; traces of albumen. H Color of urine weaker; no albumen. 9 Almost normal-colored urine; no albumen; dog is well and 10 liwly. Nothing abnormal. Total.. 4 COMPILED DATA UNDER GREEN TABLE NUMBERS. No. 2. [1 gram = 217 mg = 152 grains.] 103 Date. Dose. Weight. 24 hours' urine. General condition of animal and urine. 1902. Jan. 25 26 Grams. 2 Kilos. 4.6 cc. 200 Soon after giving color heavy vomiting and diarrhea: dog does not eat or drink the rest of the day; urine acid; no albumen. Vomiting and diarrhea continued; dog is tired and run down, 27 and began to eat; urine dark brown and no albumen. Urine same; dog livelier. 28 29 2 2 4.3 180 200 Diarrhea and vomiting stopped; urine of brown color; acid; no albumen. Urine orange; no albumen; acid. Total.. 6 G. T. 102. Trade names. — Fast Red ; Roccellin ; Cerasin ; Rubidin ; Fast Red A; Rauracienne; Orcellin No. 4. Scientific name. — Sodium salt of para-sulphonaphthalene-azo-beta- naphthol. Shade. — Brownish red. Not offered. Discovered and patented. — 1877. FAVORABLE. 1. Permitted by the law of Italy. 2. Permitted by the law of Austria. 3. Cazeneuve (Arch. gen. de mtd., 1886, p. 753) says it may be taken without effect by man or animals, sick or well, in large dot 4. Cazeneuve and Leplne (Bull, de Vacad. de mcd. 1886, p. 643): Tolerated by man well or sick. G. T. 103. Trade names. — Azorubin S; Carmoisin; Azo Acid Rubin; Fast Red C; Azorubin A. Names under which it was offered on the United States market as a food color in 1907. — Claret Red RZ; Cardinal 3 B; Azorubin; Car- moisin B. Scientific name. — Sodium salt of para-sulphonaphthalene-azo- alpha-naphtbol-para-sulphonic acid. Discovered and patented. — 1883. Shade. — Red. Offered by 6 ou1 of 12 sources. I AVnli.MI! |, Permitted by Confectionera' List. Cazknkivi; am» I.i iinj : Not poJSODOUl t" human frflinm Meyek (./. Anur. Chun. S,,c. One hundred milligrams mr kilogram body weight, <n tin United State* market ■ food color in 1907. Claret Red RZ; Red; Amaranth B (Azo color similar to) ; Bordeaux S; Naphthol Red S; Amaranth. Scientific name. Sodium Bali <»f para-8ulphonaphthalene->azo-beta- naphthol-disulphonic acid (R). Discovered and /> nt<\ 7 <>ut <>f 12 souro 106 COAL-TAR COLORS USED IN FOOD PRODUCTS. FAVORABLE. 1. Cazeneuve and Lepine (see Weyl, p. 115): Not poisonous to human beings. 3. Lieber (p. 148): A rabbit received 284 milligrams per kilogram body weight, or 199 grains per 100 pounds once a day, five times every other day, and aside from decreased appetite the second and third day of the observation period nothing untoward is noted. 4 Weyl {p. 31): "* * * not poisonous to human beings and dogs, * * * Purple * * *" 5. Cazeneuve and Lepine (Bull de Vacad. de mid., 1886, p. 648): Tolerated by man, sick or well. G. T. 138. Tradenames. — Fast Brown G; Acid Brown. Scientific name. — Sodium salt of bi-sulphobenzene-disazo-alpha- naphthol. Discovered. — 1882. Shade. — Brown. Not offered. FAVORABLE. 1. Weyl (p. 134): "The following is a summary of the results obtained with the Dis- azo colors submitted to test by me, viz, Fast Brown G * * *. All these proved to be nonpoisonous * * *." A. A dog weighing 9.63 kilos received 312 milligrams per kilogram body weight, or 218 grains per 100 pounds body weight; the same dose was repeated 48 hours afterwards, when diarrhea set in, and the urine was colored red ; 24 hours afterwards 208 milligrams per kilogram body weight, or 146 grains per 100 pounds body weight were given; the urine was colored strongly red 24 hours thereafter. Eight days afterwards 520 milli- grams per kilogram body weight, or 364 grains per 100 pounds body weight, were administered ; marked diarrhea set in; 3 days later the dose given was twice the dose last given, when severe diarrhea resulted, but unchanged color appeared in the urine, and continued for 24 hours afterwards ; diarrhea continued for 96 hours. Evidences of albuminuria apparently not dependable. B. A second dog, weighing 5.9 kilos, received 339 milligrams per kilogram body weight, or 237 grains per 100 pounds body weight daily throughout an entire month; diarrhea was produced after 6 days, which continued almost during the month; appetite was diminished and the loss in weight was about 20 per cent on the original. C. Weyl concludes as follows: "According to these experiments, this color in continuous, though slight, doses, or in large doses, but less frequently, produces diarrhea, anorexia, and emaciation." (p. 136.) unfavorable. 1. Prohibited by Confectioners' List. 2. Weyl (p. 147): "Of the remaining colors * * * others (produce) diarrhea (Fast Brown * * *)." G. T. 160. Trade name. — Crocein Scarlet 3 B; Ponceau 4 RB. Scientific name. — Sodium salt of sulphobenzene-azo-benzene-azo- beta-naphthol-monoeulphonic add. Discovered and yatented. — 1881. Shade. — Scarlet. Not offered. COMPILED DATA UNDER GREEN TABLE NUMBERS. 107 FAVORABLE. Nothing. UNFAVORABLE. 1. Prohibited by Confectioners' List. DOUBTFUL. 1. Houghton (/. Amer. Chem. Soc, 1907, v. 29, pp. 1351-57): Hinders fibrin diges- tion at all strengths; at 1: 200 hinders casein and albumen digestion. G. T. 163. Tradenames. — Biebrich Scarlet; Ponceau B; New Red L; Ponceau 3 RB; Fast Ponceau B; Imperial Scarlet. Scientific name. — Sodium salt of sulphobenzene-azo-sulpho-ben- zene-azo-beta-naphthol. Discovered. — 1878. Shade. — Scarlet. Not offered. FAVORABLE. 1. Permitted by law in Italy. (See Licber, pp. 18, 22, 23.) UNFAVORABLE. 1. Prohibited by Confectioners' List. G. T. 164. Trade name. — Crocein Scarlet O extra. Scientific name. — Sodium salt of sulphobenzene-azo-sulphoben- zene-azo-beta-naphthol-sulphonic acid. Discovered. — 1888. Shade. — Scarlet. Not offered. FAVORABLE. Nothing. UNFAVORABLE. 1. Prohibited by Confectioners' List. G. T. 166. Trade name. — Wool Black. Scientific name. — Sodium salt of sulpliobenzeiu'-uzo-sulphohen- zene-azo-para-tolvl-bcta-naplitliylainin. Discovered. — 1885. Shade. — Black. Not offered. I \\ "KAMI 1 1. Permitted by ( Sonlectioners' List 2. Wkyi. (y. //;■: "The following Is a summary "f the reeulti obtained with the Disa/o colon submitted to teet l>y me, viz: * * * Wool Back * * *. All thete proVrd to h«- IloM])(>isonOU8 * * *." 108 COAL-TAR COLORS USED IX FOOD PRODUCTS. 3. A. A dog received 167 milligrams per kilogram body weight, or 117 grains per 100 pounds body weight each day for 3 successive days; no color was admin- istered for the next 2 days, and the third day the dose was double the former dose. The urine was colored bluish-black and contained albumen. Two days afterward the same dose was given; urine of intense dark-blue color, and con- tained unaltered coloring matter, which disappeared in 48 hours. Albuminuria continued for about 15 days. (p. 137.) 13. "Wool Black is nonpoisonous both by gastric and by subcutaneous administration." (p. 137.) G. T. 169. Trade names. — Crocein Scarlet 7 B; Ponceau 6 RB; Crocein Scarlet 8 B. Xames under which it was offered on the United States market as a food-color in 1907. — Sodium salt of sulphotoluene-azo-toluene-azo- beta-naphtliol-alpha-sulphonic acid . Discovered and patented. — 1881. Shade. — Red. Offered by 1 out of 12 sources. FAVORABLE. 1. Permitted by Confectioners' List. G. T. 188. Trade names. — Naphthol Black B; Brilliant Black B. Name under which it was offered on the United States market as a food- color in 1907 — Naphthol Black BDF. Scientific name. — Sodium salt of disulpho-beta-naphthalene-azo- alpha-naphthalene-azo-beta-naphthol-disulphonic acid. Discovered and patented. — 1885. Shade. — Black. Offered by 1 out of 12 sources. FAVORABLE. 1. Permitted by Confectioners' Li. tS9)\ Where a dog, receiving subcutaneously 81 milligrams per kilogram body weight, or 22 grains per LOO pounds body weight] subcutaneously, died apparently wholly us :i result of the color. COMPILED DATA UNDER GREEN TABLE NUMBERS. 109 2. "Naphthol Black P, however, is plainly poisonous when introduced into the sub- cutaneous cellular tissue." (p. 147.) G. T. 197. Trade names. — Bismarck Brown; Phenylene Brown; Leather Brown; English Brown; Manchester Brown; Vesuvin; Cinnamon Brown. Names under which it was offered on the United States market as a food-color in 1 907 . — Vesuvin 4B Cone. Z; Bismarck Brown; Bismarck Brown B No. 216; Bismarck Brown B. Scientific name. — Hydrochloric 1 of benzene-disazo-phenylene^liamin, Discovered and patented. — 1863. Shade. — Reddish brown. Offered by 4 out of 12 sources. FAVORABLE. 1. Weyl (p. 115): " Nonpoisonous Bismarck Brown." IN! AVIIHAHI V . 1. Prohibited by Confectioners' List. 2. Wbtl (p. 117): A. A dog received 33 milligrams per kilogram body weight, or 23 grains per hundred pounds body weight ; in one and one-half hours, vomiting; next day BBme dose same result with the addition that the animal took no food and moved about but little for 48 hours; on the fourth day same dose, in two hour-, vomiting; for 9G hours animal took no food; on the eighth day albumen in the urine and the animal ate; on the ninth day 83 milligrams per ki! body weight or 58 grains per 100 pounds body weight, vomited for one-half hour after administration; for the next 5 days the animal took hardly any food; on the fourteenth day the animal improved, took food on the fifteenth day; tra albumen in urine for 17 days longer, at end of which time animal recovered. B. A dog received L69 milligrams per kilogram body weight, or L18 grains per 100 pounds body weight; in 24 hours the urine was colored brown; 48 hours after the first dose that dose was repeated, and unconverted color was found in the urine; 48 hours later the same dose was re peate d; the color in the urine disappeared in 24 hours; 4 days later the dose W8S trebled, and the animal vomited, Beemed sick for 4 days thereafter; on the fifth day reco\ cry apparently complete. On the sixth day the Last dose was repeated, and the animal vomited after tin- administration. It took QO food for 24 hours, and was normal after 48 hours. There was uo albuminuria in this ca 0. A dog received daily for an entire month 45 milligrams per kilogram bod] pht, or 31) graini per LOO pounds body weight; it was in good health during the entire time, did not vomit, and at€ M usual. Its gaio in weight I per cent . 3. "Bismarck Brown produces, when administered I the itomach, even in odv weight (or :J1J grains p»-r L00 pounds body weight), eveo when frequently admim seem to be entirely barmle t. Doses of L6 milligrams are harmless even when introduced into the subcutaneous cellular tissue " (p. tI8.) 4. "Of the remaining colon some produce vomitii . Bismarck Brown 110 COAL-TAR COLORS USED IX FOOD PRODUCTS. 5 Lewix (Lehrbuch der Toxikologie. 1897, p. 231): Produces eczema, and cites Deutsch Med. Wochenschr. t 1891, p. 45. 6. Fraexkel (p. 575): "When the Azo dyestuffs do not contain any sulpho group they are poisonous. Thus, for example, Bismarck Brown * * *. This produces no effect in email doses; on the other hand, doses of 350 milligrams per kilogram of animal (245 grains per 100 pounds) produce albuminuria and vomiting." DOUBTFUL. 1. Houghton* (/. Amer. Chem. Soc, 1907, v. 29, pp. 1351-1357): Hinders digestion of fibrin, casein and albumen, in strengths of 1:100 or 1:400. G. T. 201. Trade names. — Manchester Brown EE; Bismarck Brown R; Vesuvin B. Names under which it was offered on the Vnited States market as a food color in 1907 . — Vesuvin 4 B Cone. Z; Vesuvin B. Scientific name. — Hydrochloric! of toluene-disazo-meta-tolylene- diamin. Discovered and patented. — 1878. Shade. — Reddish brown. Offered by 2 out of 12 sources. UNFAVORABLE. 1. Lewix (Lehrbuch der Toxicologic, 1S97, p. 231): "Produces eczema," and cites Deutsch. Med. Wochenschr., 1891, p. 45. Note. — The literature is not always conclusive as between Nos. 197 and 201, and probably most, if not all, the references under No. 197 also apply to No. 201. G. T. 240. Trade name. — Congo Red. Name under which it was offered on the United States market as a food color in 1907. — Congo. Scientific name. — Sodium salt of diphenyl-disazo-binaphthionic acid. Discovered and patented. — 1884. Shade. — Red. Offered by 1 out of 12 sources. FAVORABLE. 1. Weyl: "According to Experiments 1 and 2 Congo Red is, after long-continued administration by the stomach, harml- 2. A. A dog received 274 milligrams per kilogram body weight, or 192 grains per 100 pounds ; this dose was repeated next day, when the urine became pale, was strongly- alkaline and contained albumen; the next day the same dose was repeated, whereupon the urine was of a weak red color, and a little albumen present; the next day the dose was increased to 411 milligrams per kilogram body - grains per 100 pounds; the urine was somewhat reddish and con- tained a little albumen. No color was given for 6 days, at the end of which time the urine was reddish, and deposited i reddufa sediment, probably Congo. At the end of that time the dose was increased to 685 milligrams per kilogram COMPILED DATA L'NDEE GREEN TABLE NUMBERS. Ill 2. A. — Continued. body weight, or 480 grains per 100 pounds body weight; the urine was feebly alkaline and contained some albumen. The next day the dose last given was doubled; the urine was colored to such an extent that it could be dyed with. Two days later the same dose was repeated; animal took but little food, wa3 otherwise comfortable; there was little albumen present. The loss in body weight was about A.\ per cent. (p. 141.) B. A second dog received 233 milligrams per kilogram body weight, or 163 grains per 100 pounds, daily for one month, and remained entirely well. G. T. 269. Trade name. — Chrysamin R. Xame under which it v:as offered on the United States market as a food color in 1907. — Chrysamin R. Scientific name. — Sodium salt of ditolyl-disazo-bisalicylic acid. Discovered and patented. — 1884. Shade. — Yellow. Offered by 1 out of 12 sources. FAVORABLE. 1. Permitted by Confectioners' List. 2. Weyl (p. 134 : "The following is a summary of the results obtained with Disazo colors submitted to test by me, viz: Chrysamin R." "AD of these proved to be nonpoisonous, * * *.*' 3. A. A dog received 515 milligrams per kilogram body weight, or 361 grains per 100 pounds; urine became alkaline and yellowish, and easily dyed cotton; very little albumen. Within 24 hours the same dose was repeated, diarrhea resulting and vomiting for 3 days, whereupon the animal was given one-fifth of the dose, or 103 milligrams per kilogram body weight, that is, 72 grains per 100 pounds body weight; the urine continued slightly colored, and contained a distinct amount of albumen, continuing for 2 days, when the last dose was repeated; 24 hours afterwards the dose was doubled, and repeated the next day; the day after the dose was increased 50 per cent; 3 days later the last dose was repeated; slight albuminuria set in, lasting 5 days. B. A dog received three doses of 619 milligrams each per kilogram body weight, or 433 grains per 100 pounds body weight, three times in the course of 10 days; the urine was yeUowiflh in color and contained very little albumen, (p. 145.) 4. "Chrysamin is harmless when taken into the stomach 5. Schacherl (p. : Chrysamin is harmless under the conditions in which it is used. UNFAVORABLE. 1. WlTL(p. U ^he remaining colors some produced vomiting * * ■ others diarrhea (* * * Chrv.-uniin R * * *.") G. T. 277. Trade names. — Benzopurpurin 4 B: Cotton Red 4 B; Sultan Red 4 B. 8c%efUi) ■ . — Sodium >alt of dit<>lvl-di-az<>-l>inaplitlii<>nic acid. Discovered, l^i B5. 8hadi . Red. No! offered. 112 Nothing. COAL-TAK COLORS USED IN FOOD PRODUCTS. FAVORABLE. UNFAVORABLE. 1. Chlopin {p. 130): On his own experiments classes it as suspicious. The experi- mental data are as follows: Experimental data by Chlopin. [1 gram=125 mg=87.5 grains.] General condition of animal and urine. Date. Dose. Weight. 24-hour urine. 1901. Apr. 16 17 Grams. 2 Kilos. 8 cc. 300 330 450 470 370 320 310 370 310 18 3 19 20 21 2 22 23 24 Total. 7 Dog is well, lively; urine normal color; acid; no albumen. Do. Vomiting several times; no albumen. Urine yellow-orange; no albumen; no vomiting. Do. Do. Do. Urine normal; dog is well. Do. DOUBTFUL. 1. Winogradow (Zts. Nahr. Genussm. 1903, v. 6, p. 589) says it almost completely inhibits digestion. G. T. 287. Trade name. — Azo Blue. Name under which it was offered on the United States market as a food color in 1907. — Azo Blue. Scientific name. — Sodium salt of ditolyl-disazo-bi-alpha-nuphthol- para-sulphonic acid. Discovered. — 1 885. Slvade. — Grayish violet. Offered by 1 out of 12 sources. FAVORABLE. 1. Permitted by Confectioners' List. 2. Weyl I p. tS4)i The following is a summary of the results obtained with the Disazo colors submitted to test by me, viz: * * * Azo-blue, * * * ." ''All these proved to be nonpoisonous." 3. "Azo-blue Lb harmless, both when administered by the stomach and subcuta- oeously.'i A. A dog received 237 milligrams per kilogram body weight, or 166 graina per LOO pounds; 2 days later this same dose was repeated, and the urine was a \i<»ict color strongly alkaline ami contained a little albumen. The next day tie- dose was increased to 2.5 times; colorless urine, and little albumen; there was do phenol. 'The next day I I imee the original dose was administered; abundant bluish-violet urine and little albumen. No administration for 4 <\.r. i, when 2.5 times the original dose wen given, and that d<>so repeated 3 days later; during this time ihe urine was colorless and contained little albumen; there was a very slight increase in weight. B. A dog received 319 milligrams per kilogram body weight, or 233 grains per 100 pounds daily Cor one month; animal remained well with good appetite; a slight amount of albumen made its appearance in the urine, (p. 144.) COMPILED DATA UXDER GBEKN TABLE NUMBERS. 113 UNFAVORABLE. 1. Weyl (p. 144)' "A slight amount of albumen made its appearance in the urine. " G. T. 394. Trade names. — Dinitrosoresorcin; Dark Green; Russian Green; Alsace Green; Fast Green O; Chlorin; Fast Myrtle Green. Scientific name. — Dinitroso-resorcinol (Dioximidoquinone). Discovered and patented. — 1875. Shade. — Green. Not offered. FAVORABLE. 1. Permitted by Confectioners' List. 2. Weyl (p. 63): " According to the above experiments Dinitroso-resorcinol is not dangerous to dogs when administered by the stomach even in large doses; while hypodermic administration proves fatal within 24 hours, in the proportion of 190 milligrams per kilogram of body weight" (that is, 132 grains per 100 pounds). 3. A. A dog received 173 milligrams per kilogram body weight, or 121 grains per 100 pounds; scanty dark-brown urine; the next day the same dose was repeated and urine continued dark brown, and contained trace of albumen as well as distinct reaction for iron; the next day the dose was repeated; animal remained lively; the day after that the dose was increased 50 per cent; no albumen. The 1< SB in body weight was only 4 per cent. B. A dog received 19S milligrams per kilo- gram body weight, or 139 grains per 100 pounds; seems to have been loss of appetite, coupled with dark-brown, almost black urine; 2 days later the d> - doubled, and the urine was colored green by ferrous oxid, and contained no albumcTi nor sugar; 2 days later the dose was increased 50 per cent, and some albumen was then found in the urine, (p. 62.) G. T. 398. Trade name. — Naphthol Green B. Names under which it was offered on the United States market as a food color in 1907. — Naphthol Green; Naphthol Green B. Scientific name. — Ferrous sodium salt of nitroso-betanaphthol- beta-monosulphonic acid. Discovered. — 1883. Shade. — Green. Offered by 2 out of 12 sources. • >K Mi! I l. Wsyl(ji '■ ■ \ A dog received 172 milligrams per kilogram body weight, or 120 graini i"-r loo pounds body-weight; the arise wat greenish, and conjunctiva stained intensely green; repeated -\ successive days, and the next day the dose was in cre a se d to fivefold; appetite undisturbed, and animal remained lively. B. A dog received 417 milligrams per kilogram body weight, grains per 100 pounds; there was • dirty yellowish-green color to the urine; do albumen, and OOi more than tnces Ol iron. 'The animal was normal within a day, and two days after the same repeated; do change in animal was recorded, except that on tin-day following the feces were normal, but colored • en I] Bull. L47 -12 8 114 COAL-TAR COLORS L T SED IN FOOD PRODUCTS. 2. Weyl (p. 65): "Experiments 1 and 2, in which 2 to 5 grams of the color were intro- duced directly into the stomach, demonstrated its harmlessness in this method of administration." 3. Buss lists it as nonpoisonous. UNFAVORABLE. 1. Prohibited by Confectioners' List. 2. Weyl (p. 65): "On the other hand, in a hypodermic administration, in two out of three cases abscesses and septic fever were induced." G. T. 399. Trade names. — Sun Yellow; Curcumin S; Jaime Soleil; Maize. Scientific name. — Sodium salt of the so-called Azoxy-stilbene-disul- phonic acid. Discovered. — 1883. Shade. — Yellow. Not offered. FAVORABLE. 1. Meyer (J. Amer. Chem. Soc, 1907, v. 29, p. 897): A dog received 100 milligrams per kilogram body weight, or 70 grains per 100 pounds, increased geometrically through the fourth day, when diarrhea set in; up to this time the animal had been given 19.27 grams, or 1,465 milligrams per kilogram body weight, equiv- alent to 1,026 grains per 100 pounds body weight; the average dose per day would have been 366 milligrams per kilogram body weight, or 256 grains per 100 pounds; the animal was given its fifth portion of coloring matter the same size as the fourth, thereupon color- was omitted, and for the following 7 days the dosage of the third day, which amounted to 400 milligrams per kilogram body weight, or 280 grains per 100 pounds body weight, was given; the urine was col- ored orange throughout the entire test after the first day; the fecal matter also was of orange color; slight diarrhea on the fourth and twelfth days of the test, and vomiting on the fifth day, the cause of which does not seem to have been definitely determined; the autopsy revealed nothing abnormal. G. T. 425. Trade names. — Auramin; Auramin O; Pyoctanin Aureum. Names under which it was offered on the United States market as a food color in 1907. — Auramin O; Auramin; Canary Yeilow. Scientific name. — Ilydrochlorid of Lmido-tetramethyl-diamido- diphenylmethane. Discovered. — 1883 . Shade. — Greenish yellow. Offered by '•*> out of L2 sources. Nothing. I \\ OB A HI I . i NFAVORAB1 I I. Chlopin (p. tS7): On his own experiments classes i1 m poisonous. The experi- mental data are aa follows: COMPILED DATA UNDER GREEN TABLE NUMBERS. Experimental data by Chlopin. [1 gram =09 mg=48 grains.] 115 Date. Dose. Weight. 24-hours' urine. General condition of animal and urine. 1901. Mar. 2-4 6 Grams. 1.0 Kilos. 14.4 cc. 550 340 Dog quite normal; acid; no albumen. Urine strong brown yellow; traces of albumen; appetite less. Thin stool. 7 8 Diarrhea; urine brown yellow, pales with sulphuric acid, and contains much albumen. Stool normal; color of urine weaker; much albumen; eats well; 9 10 lively. Stool normal; urine less colored; less albumen. 11 335 Stool normal; urine less colored; traces of albumen. 12 Stool normal; urine less colored; insignificant traces of albumen. 13 Do. 14 2.0 13.2 Do. 15 Strongly brown yellow; much albumen. Dog depressed; lies down; eats little. Vomits and diarrhea; urine strong yellow; little albumen; takes only milk. Vomits and diarrhea; urine strong vel low; much albumen. 16 None. 295 480 17 19 20 Feces normal color; color urine almost normal; much albumen. 21 352 380 Feces normal color; color urine almost normal; albumen less. 24 10.4 13.6 Do. 1 Subcutaneously. DOUBTFUL. 1. Winogradow (Zts. Xahr. Genussm., 1903, v. 6, p. 589) says it noticeably retards digestive action; is not indifferent. G. T. 427. Trade names. — Malachite Green; New Green; Fast Green; Benzal Green; Diamond Green B; Malachite Green B; New Victoria Green; Vert Diamant; Bitter-almond-oil Green. Names under which it was offered on the United States market as a food color in 1907. — Green M; New Green Crystals; Green 088. Scientific name. — Zinc double-chlorid, oxalate, ferric double-chlorid of tetramethyldi-para-amido-triphenyl-carbinol. Discovered. — 1877-78. Shade. — Bluish green. Offered by 2 out. of 12 sources. I AVOKAI1I.K. 1. Permitted by Confectioners' List. 2. W'i^i. "According i<> Gmndhomme * * * Malachite Green are | is) ;il-<> Qonpoisonous." 3. "* * * Malachite Green are (is) as is now established, almost without poison' <>ii- action." (p. 55.) •i. I.i ui\ {Lehrbuch (It 'V<>r'ik<>h><; ays when free from arsenic it is harmless. ."> Buss list M ;i- iionp ONI 4VORAB1 i 1 l'i\/ i .1' - r,„r. path, phnrm., ! ( >n<« hundred in ill i- grams per kilogram body weight o! rabbit, or 70 grams per 100 pounds, injected subcutaneously, caused after the third day motor paralysifl and cramps, which resulted fatally at tin- end of the ninth day. 116 COAL-TAR COLORS 1 T SED IN FOOD PRODUCTS. 2. Lewin (Lehrbuch der Toxihologie, 1897, p. 231): In the case of one workman, in contrast with others who had long been unaffected by this substance, itching, burning, inflammation, and swelling of hands and feet, and formation of blisters occurred. G. T. 428. Trade names. — Brilliant Green; New Victoria Green; Emerald Green; Malachite Green B; Ethyl Green; Fast Green J. Names under which it was offered on the United States market as a food color in 1907 . — Green E; Green 087; Emerald Green Crystals. Scientific name. — Sulphate of zinc-double-chlorid (rarely oxalate) of tetraethyl-diamido-triphenyl-carbinol. Discovered.— 1879-80. Shade. — Yellowish green. Offered by 3 out of 12 sources. FAVORABLE. 1. Lewin (Lehrbuch der Toxikologie, 1897, p. 231) says when free from arsenic it is harmless. 2. Buss lists it as nonpoisonous. UNFAVORABLE. 1. Chlopin (pp. 171-2): Classifies it as "very poisonous" on his own experiments (see p. 181). The experimental data are as follows: Experimental data by Chlopin. No. 1. [1 gram=125 mg=87.5 grains.] Date. Dose. Weight. 24 hours' urine. General condition of animal and urine. 1901. Apr. 24 25 Grams. Kilos. 8 cc. 410 Before experiment dog is well; urine normal. 2 Soon after giving dye vomiting and diarrhea; in the evening 26 395 only drank water; vomiting kept up. No vomiting; urine greenish; acid; insignificant traces of 27 albumen. Do. 28 2 400 No vomiting; urine greenish; acid; no albumen. 29 Do. 30 395 Do. May 1 2 Urine normal; dog quite well. Do. 3 395 Do. Total.... 4 No. 2. 1 1 gram -=111 mg=7« grains.) 1902. Jan. 21 23-24 1 9 400 L'"» 2 7.2 5 Before experiment dog quite normal; urine normal; after fir- ing dye very violent vomiting several times. i i in.- greenish In color; add; vary mudb albumen; no vomiting. Died durta i the ol COMPILED DATA UNDER GREEN TABLE NUMBERS. 117 Experimental data by Chlopin — Continued. No. 3. [1 gram=119 mg=83 grains.] Date. Dose. Weight 24 hours' urine. General condition of animal and urine. 1902. Jan. 28 30 31 Total... Grams. Kilos. 8.4 Before experiment dog and urine normal; soon after giving dye vomiting began, lasting over an hour; dog stands wiih diificulty. During ni^ht dog improved somewhat, began to eat; drinks much; soon after giving dye vomiting. During night vomiting and diarrhea; during night 30th, 31st, in bad condition; does not take food. During night 31st second dog found dead in cage. Cause of death, " paralysis of the heart." 2. Lewix (Lehrbuch der Toxikologie, 1897, p. 231): In the case of one workman, in contrast with others who had long been unaffected by this substance, itching, burning, inflammation and swelling of hands and feet, and formation of blisters occurred. G. T. 433. Trade name. — Guinea Green. Name under which it was offered on the United States market as a food color in 1907. — Guinea Green B. Scientific name. — Sodium salt of diethyldibenzyl-diamido-tri- phenyl-carbinol-disulphonic acid. Discovered. — 1883. Shade. — Green. Offered by 1 out of 12 sources. FAVORABLE. 1. Chlopin (p. 174): On his own experiments classified it a* experimental data are as follows: Experimental data by Chlopin. [1 gram=200 mg= 140 grains.] nonpoisonous. The Date. Dose. Weight 24 hours' urine. 1901. May 18 19 Grams. 2 Kilos. 5 cc. 2. r »0 2 16 270 270 280 20 21 a 2 2 si 10 General condition of animal and mine. Dog and urine normal before experiment aton, no albumen; vomited el night. No vomiting. Urine slightly greenish; no albumen. Urine normal In color and composition. Do. Da Do. I)... Do. Do, Da G. T. 434. Track names.— Light Greeo SF bluish; Acid Green. Scientific name. Sodium Ball of dimethyldibenxyl-diainido-tri" phenyl-carbinol-1 risulphonic acid. 118 COAL-TAR COLORS USED HSf FOOD PRODUCTS. Discovered. — 1879. Shade. — Green. Offered by 1 out of 12 sources. Nothing. FAVORABLE. UNFAVORABLE. 1. Chlopin (pp. 176-7): Examined this color, and on his own experiments classes it as "nonpoisonous, but not entirely indifferent." The experimental data are as follows : Experimental data by Chlopin. No. 1. [1 gram=133 mg=93 grains.] Date. Dose. Weight. 24 hours' urine. General condition of animal and urine. 1902. Apr. 24 25 Grams. 2 Kilos. 7.5 cc. 380 Dog and urine normal. Urine green; acid; no albumen. 26 347 297 Urine green; traces of albumen. 27 Do. 28 2 Do. 29 Urine green; no albumen. 30 360 Urine less green; no albumen. 31 Do. May 1-3 Do. 370 Do. Total . 4 No. 2. [1 gram=110 mg=77 grains.] 1902. Aug. 5 6 3 9.1 300 310 Dog and urine normal. Urine greenish, acid, no albumen. 7 3 3 3 3 Do. 8 320 350 Do. 9 Do. 10 Do. 11 400 350 Do. 12 9 Do. Total . 15 DOUBTFUL. 1. Winogradow (Zts. Ndhr. Genussm. 1903, v. 6, p. 689) says it noticeably retards digestive action; is not indifferent G. T. 435. Trade names. — Light Green SF yellowish; Acid Green; Acid Green extra cone. Names under which it was offered on th< United Shifts market as a food color in 1907. — A.cid Green cone. V X; Light Green SF yellow shade; Acid Green cone. 780 ; Pistachio. COMPILED DATA UNDER GREEN TABLE NUMBERS. 119 Scientific name. — Sodium salt of diethyldibenzyl-diamido-tri- phenyl-carbinol-trisulphonic acid. Discovered. — 1879. Shade. — Green. Offered by 4 out of 12 sources. FAVORABLE. 1. Lieber (p. 144): The animal was a fully developed male guinea pig, and received per kilogram body weight 240 milligrams, or 108 grains per 100 pounds, live times in all, every other day. There was apparently nothing irregular or ab- normal observed during the whole test of nine days. G. T. 448. Tradenames. — Magenta; Fuchsin; Rosein; Anilin Red. Obsolete names. — Rubin ; Solf erino ; Fuchsiacin ; Rubianite ; Azalein ; Erythrobenzin ; Harmalin. Names under which it was offered on the United States market as a food color in 1907. — Magenta powder A; Fuchsin Crystals; Magenta FABSRed 101. Scientific name. — Mixture of hydrochlorid or acetate of pararo- sanilin (triamidotriphenylcarbinol) and rosanilin (triamidodiplienyl- tolylcarbinol) . Shade. — Bluish-red. Offered by 4 out of 12 sources. FAVORABLE. 1. Permitted by Confectioners' List. 2. Weyl (p. 22): "The colors examined * * * Fuchsin were (was) found to be nonpoisonous-;" "Similarly a hen which had eaten for three weeks oate covered with fuchsin was in good health." (p. 24.) 3. "According to Grandhomme rabbits bear without injury fuchsin free from arsenic * * *." (p. 31.) 4. "Fuchsin * * * (is) as is now established, almost without poisonous action." (j>. 55.) 5. Fkaenkel (p. 574), quoting Penzoldl. says that it Lb entirely nonpoisonous, and completely prevents putrefaction. <;. Permitted by the law of Austria. 7. Lkwin {Lehrbuch dar Toxtkologie, 189 0), says when free from arsenic it is harmless. 8. Clouet am. Bbrgeron (./. pharm. chim., I871 t i . I me of them took •ually 500 milligrams, that i-. 7.7 grains in L6 days; there was no digestive disturbance of any kind, and the urine, which was examined daily, contained no albumen. They cite a case of Blight's disease, in which the amount of albumen decreased when fuchsia was administered, and they conclude that fuchsin may be good Corsui no Blight's disea ( KFAVOBABl I , 1. Forbidden by the law of Fran , p. 81.) 2 Chlopim (] in lined this color, and on hi- own experiments classes it as "Sn acauee "i" vomiting and tracee of albumen." 'The experimental data are ;i ~ foU 120 COAL-TAR COLORS USED IX FOOD PRODUCTS. Experimental data by Chlopin. Date. Dose. Weight. 24 hours' urine. General condition of animal and urine. 1901. Apr. 18 Grams. 3 Kilos. cc. 270 280 Dog before experiment well and urine normal; vomited several 19 times after receiving dye. Urine quite red; acid; no albumen; general condition normal. 20 Do. 21 2 320 22 Urine darkish; acid; traces albumen. 23 Color normal; no albumen. 24 Do. Total. Note.— This sample may have contained some phosphin, G. T. 532, see page 133. DOUBTFUL. 1. Wixogradow (Zts. Nahr. Genussm. 1903, v. 6, p. 589) says it almost completely inhibits digestion. G. T. 450. Trade names. — Hofmann Violet; Dahlia; Red Violet 5R extra; Violet R; Iodin Violet; Primula; Violet 5 R; Violet R R. Scientific name. — Mixture of the hydrochlorids or acetates of the monodi- or trimethyl- (or ethyl-) rosanilins and pararosanilins. Discovered. — 1863. Shade. — Violet. Not offered. FAVORABLE. 1. W'EYh(p.H): "* * * Anilin Violet (Dahlia) 2. Buss lists it as nonpoisonous. * ■* •* (is) also nonpoisonous. ' * IM AVOUABLE. 1. Fraenkel (p. 574) quotes Penzoldt, and says it completely arrests development, and causes muscular paralysis. G. T. 451. Trade names. — Methyl Violet B; Direct Violet; Dahlia; Paris Violet; Violet de Methylanilin ; Pyoctanin. Names under which it was offered on the United, States market as a food color in 1907.— Methyl Violet; Methyl Violc! B; Methyl Violet BB extra; Methyl Violet 3 BD. Scientific name. — Hydrochloric! of penta- and bexamethyl- para- rosanilin. Discovered. — 186 1 . Shade. — Violet. Offered by 5 out of L2 sources. I \\ ORABLB. 1. Permitted by Confectioners' List. 2. Permitted i>y the Austrian law. COMPILED DATA UNDER GREEX TABLE NUMBERS. 121 3. Weyl (p. 24): " * * * Anilin Violet (Dahlia) * * * [is] also non- poisonous." 4. " * * * Methyl Violet [is] as is now established, without poisonous action. " (p. 5S). 5. Fraexkel (p. 573): "Methyl Violet * * * is relatively n on poisonous." 6. Buss lists it as nonpoisonous. UNFAVORABLE. 1. Graefe and Braunschweig (Fortschr. Medizin, 1890, i. 8 t p. 405): "It seems to be proven that damage will actually result even in the case of most cautious use, which we are sure we exercised." 2. Santori (MokschotVs Untersuchungen, 1895, v. 15, p. 52): I. A dog weighing 7,600 grams received 5.6 grams dye in 12 days; this amounts to 61 milligrams per kilo per day, or 43 grains per 100 pounds per day. Continued vomiting beginning with 0.1 gram dye; progressive emaciation and general falling away; catarrh of eyes and nose; distinct dislike for food and great desire to sleep; temperature below normal, urine unchanged. Loss of weight 1.000 grams, or 21 per cent. The animal died on the thirteenth day. The autopsy showed a pale and blood-poor liver; the kidneys were in a typically congested condition and contained accumulations of blood corpuscles. II. A second dog weighing 6,000 grams received 3.8 grams dye in 14 days, which amounts to 71.4 milli- grams per kilo per day, or 50 grains per 100 pounds per day. The animal died on the fourteenth day. There was daily vomiting and rapid emaciation; final weight loss was 1,100 grams, or 18.3 per cent; temperature normal; bloody urine beginning the eighth day. The autopsy showed a blood-poor liver, soft and swollen epithelia; kidneys the same as in the case of the preceding dog. G. T. 457. Trade names.— Anilin Blue, spirit-soluble; Spirit Blue; Fine Blue; Bleu Lumiere ; Opal Blue; Gentian Blue 6B; Hessian Blue; Bleu-de- Nuit. Scientific name. — Hydrochloric! sulphate or acetate of triphcnvlro- sanilin and triphcnylparaiosanilin. Discovered. — 1860-1862. Shade. — Greenish blue; not offered. 1 AVORAIJI.K. 1. Permitted by Confectioners' List. LNTOR] \foU9ChoU*i ( ' nU rtUckli mjt n , IS'.).',, r. 15, /'. it is har: A dog wei In: •• or B8 grains per LOO pounds per day. Weight remained the same, genera] condition good, urine and temperature unchanged; killed by chloroform; autopsy showed everything normal. 3. Lbbbb (p. / 1 . where it u stated to be permitted by the Austrian law i pp. jj-j.; , where it i- stated to !>»• permitted by the Italian law {p. 31 , where it is stated to be permitted by the French law in candies, pastilles, sweetmeats, sauces, fruit-, ami certain liqueur- ordinarily not colored. 4. Wiyi. . ..'.'), quoting Sonnenkalb (p. t4), quoting Grandhomme. iaj \k! 1. 1 (p. I that it it effective in only 5 percent of malaria cases. 6. Permitted by the law of Austria. 7. l'.i si lists it as ii mpoi onous. 122 COAL-TAR COLORS USED IN FOOD PRODUCTS. UNFAVORABLE. 1. Weyl (p. 23), quoting Friedrich, where poisoning was produced in a young man engaged in packing this dye. G. T. 459. Trade names. — Iodin Green; Pomona Green; Night Green; Vert Lumiere. Scientific name. — Zinc-double-chlorid of heptamethyl-rosanilin- chlorid. Discovered and patented. — 1866. Shade. — Green. Not offered. Nothing. FAVORABLE. UNFAVORABLE. 1. Chlopin (p. 175) on his own experiments classes it as "suspicious." (See p. 181). The experimental data are as follows: Experimental data by Chlopin. [1 gram=lG7 mg=117 grains.] Date. Dose. Weight. 24 hours' urine. General condition of animal and urine. 1901. May 9 10 Grams. G Kilos. G cc. 420 470 441 420 390 400 442 370 Before experiment dog and urine normal. Do. 11 Ufine slightly greenish; no albumen. Do. 12 2 13 Urine has greenish opalescence; traces of albumen. 14 Do. 15 Urine has greenish opalescence; no albumen. Do. 16 2 17 Urine has greenish opalescence; traces of albumen. 18 400 Normal color; no albumen. Total.. G 2. Buss lists it as poisonous. DOIJHTFUL. 1. Winogradow (ZtS. NcJir. Gcnussm., 1903, V. 6, p. 689) says it noticeably retards digestive action; is not indifferent. G. T. 462. Trade names. — Acid Magenta; Acid Fuchsin; Acid Rubin; Fuch- sin S; Acid Rosein ; Rubin S. Names under which it was offered on (he United States market as a food color in t907.- Acid Magenta Powdered; Acid Magenta. Scientific name. — Mixture of the sodium or ammonium salts of the trisulphonic acids of rosanilin and pararosanilin, Discovered. L877. Shade. Red. Offered by 2 out of 12 sources. COMPILED DATA UNDER GREEN TABLE NUMBERS. 123 FAVORABLE. 1. Permitted by Confectioners' List. 2. Cazeneuve (Arch. gen. mid., 1886, p. 753) says it may be taken without effect by man and animals, sick or well, in large doses. 3. Weyl (p. 55), where he says that it is established of this color that it is almost with- out poisonous action. 2. Lieber (p. 14), where it is stated to be permitted by the law of Austria (p. 31), where it is stated to be permitted by the law of France for confectionery, cor- dials, and the like. 5. Permitted by the law of Austria. 6. Permitted by the law of Italy. 7. Cazeneuve and Lefine (Compt. rend., 1885, v. 101, p. 1011): A. Dog: 15 kilos weight, received — Days. Grains. Milligrams per kilo. Grains per 100 pounds. 15 5 5 5 1 2 5 10 67 134 335 G70 47 94 235 470 No diarrhea; no vomiting; no albuminuria; urine colored only occasionally, but did contain the leuco compound of the dye. B. Man: 1. Afflicted with Bright's disease; took two grams daily for one week; no effect. 2. Afflicted with renal cirrhosis; four grams daily for several days; no effect. 3. A well man took four grams daily for several days; no effect. 8. Cazeneuve and Lefine (Bull, de Vacad. de mid. 1886, p. 643): Tolerated by man, sick or well. G. T. 467. Trade name. — Acid Violet 6B. Scientific name. — Sodium salt of dimethyl diethyl dihenzyl triamido tri phenyl carbinol disulphonic acid. Discovered and patented. — 1889. Shade. — Violet. Not offered. FAVORABLE. 1. Santori (MoleschotCs Untersuchungen, 1895, r. 15, p. 54)'. A dog weighing 4,260 grama received 12.5 grama dye in 28 days, which amounts to LOB milligrams per kilo per day, or 74 grains per 100 pounds per day. The dog vomited the Color only two or three times and was otherwise normal. Appetite, temperature, and urine all remained normal. Conclusion: Nonpoisonous; autopsy also showed everything normal. Nothing. [JNPAVORABL] G. T. 477. Tradi names.-— Alkali Blue; Nicholson Blue; Past Blue. Scientific /num. -Mixture of Bodiuxn aaltfl of triphenyl rosanilin monoeulphonic acid and of triphenyl para roBanilin-monosulphonio acid. 124 COAL-TAE COLORS USED IN FOOD PRODUCTS. Discovered and patented. — 1862. Shade. — Blue. Xot offered. FAVORABLE. 1. Santori (MoleschotVs Untersuchungen, 1895, i\ 15, p. 45): A dog weighing 4,500 grams received 25 grams dye in 30 days, which amounts to 185 milligrams per kilo per day, or 129.5 grains per 100 pounds per day. The urine remained of normal color, but the stool was a deep blue black. Throughout the whole time the animal was in perfect health. Killed with chloroform; autopsy showed everything normal. UNFAVORABLE. Nothing. G. T. 478. Trade names. — Bavarian Blue DSF; Methyl Blue, water-soluble; Navy Blue B; Methyl Blue for silk MLB. Scientific name. — Sodium salt of triphenyl pararosanilin di- and tri-sulpho acid. Discovered. — 1862; not patented. Shade. — Blue. Not offered. Nothing. FAVORABLE. UNFAVORABLE. 1. Chlopin (p. 170) examined this color, and on his own experiments classed it "nonpoisonous but not quite indifferent." Experimental data by Chlopin. [1 gram=115mg=Sl grains.] Date. Dose. Weight. 24 hours' urine. General condition of animal and urine. 1901. May 18 19 Grams. 2 Kilos. 8.7 450 | Dog and .urine quite normal. 23 380 370 Do. 24 2 Do. 25 Do. 2 Do. 27 370 320 Do. 28 2 Do. 2'J Do. 30 2 Urine slightly greenish; traces of albumen: acid. Do. 31 June 1 Do. 2 327 Do. 3 Do. 4 3 Do. :. [Trine slightly greenish; traoas of albumen; acid; diarrhea. t. 3 105 Do. 7 i rine slightly greenish; traces of albumen; add. s 3 M6 868 870 Da /■ Toxikologie, 1897, p. 151 : RosolicAcid Is positively nan- poisonous. COMPILED DATA UNDER GREEN TABLE NUMBERS. 127 UNFAVORABLE. 1. Chlopin (p. 167) examined this color, and on his own experiments classes it as "strongly poisonous." The experimental data are as follows: [1 gram=137 mg=96 grains.] Date. Dose. w^hja* General condition of animal and urine. 1902. Mar. 12 13 Grams. 1 2 3 3 Kilos. 7.3 cc. 330 375 300 Before experiment dog quite well; urine normal color; acid; no albumen. Diarrhea; urine faint red. 14 Diarrhea; lassitude: eats little; urine red; no albumen. 15 Violent diarrhea and vomiting; dog stands on feet with diffi- culty; eats nothing. Same conditions as on preceding day; dog sick for a long time after. 16 Total.. 9 2. Prohibited by the German law of 1887. 3. Prohibited by the Belgian law of June 17, 1891. 4. Buss lists it as poisonous. DOUBTFUL. 1. Lewin (Lehrbuch der Toj Otologic, 1897, p. 231): Rosolic Acid is positively non- poisonous. Small animals can take 1 gram and more of it. In Austria it is prohibited for use in coloring foods. Corallins, or red (Pa?onin) or yellow colors, consisting of Aurin and Rosolic Acid were regarded as poisonous because in experiments on man and animals illness occurred, but are said to be poisonous only in the presence of arsenic, phenol, or anilin. These sub stances are prohibited in the coloring of food. G. T. 488 or 490. Trade name. — Victoria Blue; Victoria Blue B; Victoria Blue 4R. Scientific names. — Hydrochloric! of phenyltetra (penta) uiethvl- triamido-diphenyl-al])lia-naj)htliyl-carbinol (note: 4K is bracketed). Discovered and patented. — 1883. Shade. — Blue. Not offered. Nothing. I A\ uUAHLE. UNFAVOlIAHl .!•: chott'i Untersuchungen, 1895, v. i~>, p. 47)\ A dog weighing gran una dye in 22 days, which amounts to 45 milligram! per kilo per day or 32 grains per 100 pounds per day. Alter receiving L.5 grama the annual Buffered copious continuous Balivation, anemia, and emacia- tion and occasional vomiting. The urine's color did not change; th- ine blue; temperature slightly above and below aormal; although highly emaciated (loss in weight wi imsorSS] per cent) the animal retained fa appetite; animal died on the morning of the 23d day. The autopsy showed phagus, stomach, and intestines colored deep blue and filled with ■ green- ish scum, extended and strong catarrh of the stomach and intestines; kidneys and liver contained very little blood. Conclusion: Poison 1 1 : It is uncertain which of these two dyes Santori used 128 COAL-TAR COLORS USED IX FOOD PRODUCTS. G. T. 502. Trade name. — Rhodamin G and G extra. Scientific name. — Chiefly Triethylrhodamin. Discovered and patented. — 1891. Shade. — Bluish. Offered by 2 out of 12 sources. Nothing. FAVORABLE. UNFAVORABLE. 1. Chlopin (p. 184) examined this color, and on his own experiments classes it as "suspicious." The experimental data are as follows: Experimental data by Chlopin. fl gram= 167 mg= 117 grains.] Date. Dose. Weight. 24 hours' urine. General condition of animal and urine. 1903. Feb. 18 19 Grams. 3 3 Kilos. 6 cc. 310 Dog normal; urine acid; no albumen; after 2 hours thin stool; remainder of day lively; good appetite. Urine fuchsin color, acid, no albumen; stool and appetite normal. 20 3 3 3 Goes to stool without results; otherwise as above. 21 270 320 Do. 22 Do. 23 Goes to stool without results; urine normal after 8 days. Total.. 15 G. T. 504. Trade names. — Rhodamin B; Rhodamin O; Safranilin. Names under which it was offered on the United States market as a food color in 1907. — Rhodamin B extra; Rhodamin; Rhodamin B. Scientific name. — Hydrochlorid of diethylmeta-amido-phenol- phthalein. Discovered and patented. — 1887. Shade. — Bluish red. Offered by 5 out of 12 sources. FAVORABLE. 1. Lieber (p. 141): A young female rabbit received 339 milligram body weight, or 237 grains per 100 pounds, five limes on alternate days. "During the whole period the animal seemed to be perfectly at ease, was lively, displayed good appetite, and gained steadily * * * •" The gain in weight was, roughly, 7 per cent. i \ I \\ i»l;.\ Hi l . l . < Ihlokm (pp. U axnined this color, and on bis o* d experiments classifies it M "not poisonous, but not entirely indifferent; suspicious." The experi- mental data are as follow -: COMPILED DATA UNDER GREEX TABLE NUMBERS. Experimental data by Chlopin. f 1 gram = 109 mg = 76 grains.] 129 Date. Dose. Weight, j"*™* General condition of animal and urine. 1904. Feb. 18 19 Grams. 3 3 3 3 3 Kilo*. 9.2 cc. 420 400 400 Dog normal; urine aci; Eosin Bluish. Names under which it woe offered on tin United States market as a food color in t907. — Erythrosin Yellow Shade; Erythrosin B; Ery- throsin. COMPILED DATA UN DEB GREEN TABLE NUMBERS. 131 Scientific name. — Sodium or potassium salt of tetraiodofluoresceiru Discovered. — 1876. Shade. — Bluish red. Offered by 5 out of 12 sources. FAVORABLE. 1. Permitted by Confectioners' List. 2. Wetl • <>. ■-.' " According to Grandhomme, rabbits bear without injury * * * Erythrosine * * *." 3. Permitted by the laws of Prance 4. Permitted by the law of Austria. 5. Fbaenkel {p. 574)'- "Rose Bengal * * produces no noticeable disturb- ances." 6. Buss lists it as nonpoisonous. UNFAVORABLE. 1. Lewi\ (Lehrbuch d?r Toxikologic, 1897, p. 131)'. "The continued use of these coloring matters, as well as of phenolphthalein, which becomes colored in the system, I regard as harmful, and, in fact, through ad ion as coloring matters." G. T. 520. Trade names. — Rose Bengal; Rose Bengal A T: Rose Bengal X; Rose Bengal G. Names under which it was offered on tfa United States market as a food color in 1907. — Rose Bengal B; Phloxin B. Scientific name. — Alkaline salt of tetraiododichlorofluorescein. Discovered. -1875. Shade. — Bluish red. Offered by 2 out of 12 sources. I AVORABLE. 1. Permitted by Confectioners 5 List. G. T. 521. Troth names. — Phloxin: Phloxin TA ; Kosin 10B. Scientific name. — Sodium Bait of tetrabromotetrachlorofluorescein. Disco; red. — 1882. Shade.— Red. \<>t offered. kav«»i:\iii I l Pa mitted by < 'onfectioners' List. 2. Permitted by ili<- Austrian law. mined this color, and on bis own experiment! rliwiflra ii m "nonpoisonous. " The experimental data are m follows: 132 COAL-TAR COLORS USED IX FOOD PRODUCTS. Experimental data by Chlopin. No. l. [1 gram= 143 mg= 100 grains.] Date. Dose. Weight. 24 hours' urine. General condition of animal and urine. 1901. Apr. 10 11 Grams. 1 Kilos. 7.0 cc. 380 370 370 370 Dog and urine quite normal. Do. 12 Do. 13 3 Do. 14 Urine orange. 15 390 Do. Total.. 4 No. 2. [1 gram=116 mg=Sl grains.] 1901. Nov. 3 4 3 8.G 450 430 Dog well; urine normal color; acid; Urine orange; no albumeu. Do. Do. Do. Do. Do. Do. Color almost normal; no albumen. Do. no albumen. 5 3 6 400 440 430 450 400 7 3 3 g 9 10 3 11 12-14 Total.. 15 G. T. 527. Trade names. — Coerulein S; Alizarin Green; Anthracene Green. Scientific name. — Sodium bisulphite compound of coerulein. Discovered. — 1879. Shade. — Black. Not offered. FAVORABLE. 1. Chlopin (pp. 186-7) examined this color, and on his own experiments reports it as "non poisonous, and not sufficient data to regard it as suspicious." The exper- imental data are as follows: Experimental data by Chlopin. No. l. [1 gram ----- 117 mg— lil.'i grains.] Date. Dose. Weight. 21 hours' urine. Qeaeral condition of animal and urine. I'.HM. drum*. 2 Kiln,. ti. S cc. 820 880 800 830 800 Before experiment dog quite normal. i i me normal; acid; no albumen. Sept 1 2 8 8 Do lnn«' has scarcely perceptible sheen; m> albumen. 3 Do. 4 6 7 8 8 8 8 Do. 800 Do. Do Nothing abnormal. Do. 10 11 2 Do. Do. Do 22 COMPILED DATA UNDER GREEN TABLE NUMBERS. Experimental data by Chlopin — Continued. Xo. 2. 11 gram=lG7 mg=117 grains.] 133 Date. Dose. Weight. 2 uruiT 8 General condition of animal and urine. 1902. Mar. 3 4 Grams. 2 3 3 2 Kilos. 6 cc. 270 250 265 Before experiment dog quite well; urine normal. Urine vellow with orange sheen: acid; no albumen. 5 Thin stool twice; urine dark yellow with greenish sheen; acit offered. I ok well; urine normal. Do. 16 2 17 Do. 18 3 Urine yellower than normal; vomited several limes after taking dye; no albumen. Urine same; no vomiting. 19 330 370 21 2 22 No vomiting; urine same. Color normal. 23 ; 24 Dog well; urine normal. Total . . 7 3. (p. 178): See experimental data on G. T. 448, which also applies to this color. 4. Fraenkel (p. 578): Where its physiological action is compared with quinin its action on protozoa is far greater. "The Phosphins are locally strong irritants, and producers of inflammation of medium poisonous nature so that humans can very well bear 400 milligrams, or 6.17 grains. 5. Lewin (Lehrbuch der Toxikologic, 1897, p. 232): "Phosphin * * * produces in humans, in doses up to 1 gram, vomiting and diarrhea." DOUBTFUL. WlNOGRADOW (ZtS. Xahr. Crnimm. inhibits digestion. l'jo.). v. 6*, p. 589) says it almost completely G. T. 563. Trade names. — Alizarin Blue S; Anthracene Blue S; Alizarin Blue ABS. Scientific name. — Sodium bisulphite compound of dioxyanthra- quinonc-beta-quinolin. Discovered and patented. — 1 88 1 . SJoade. — Blue, Not offered. I Permitted bv law of Austria. FA VOKABLK. i \l \ \ ORABLE. < Ihlopin p. 171): ( m authority no1 given reports this color as poisonous or harmful. Ehrlich (Da* SauentoffbeduerfniM des Organismus, 1885, p. M): "Per kilogram of rabbit, 12- 1 5 CC of this solution in general produce death within the first quarter of an hour; whereat i cc of the same did not usually produce it, and 7 cc represent a medium, when properly applied, fatal dose." (This solution contained DOi to exceed 17 per cent, coloring matter; each cubic centimeter represents L70 milligrams per kilogram body weight, or L19 grains per 100 pounds; the coloring matter was introduced subcutaneously.) lists it as poisonous. COMPILED DATA UNDER GREEN TABLE NUMBERS. 135 G. T. 572. Tradename. — Indophenol white; Leucindophenol. Scientific name. — Tin compound of dhnethyl-para-amido-phenyl- para-oxy-alphanaphthylamin. Discovered. — 1881 . Shade. — Blue. Xot offered. FAVORABLE. Nothing. UNFAVORABLE. 1. Santori {MoleschotVs Untersuchungen, 1895, v. 15, p. 57): A dog weighing 4,000 grams received 18 grams dye in 30 days, which amounts to 150 milligrams per kilo per day or 105 grains per 100 pounds per day. Temperature, urine, and weight all remained unchanged. Animal killed with chloroform; autopsy showed fatty degeneration of the liver; everything else normal. (Santori classes this color as "not nonpoisonous.") G. T. 574. Trade names. — Ursol D; Ursol P; Ursol DD. Scientific name. — Hydrochlorids of para-phenylene diamin, para- amidophenol, and diamidodiphenylamin, respectively. Discovered. — 1888. Shade. — Brown to black. Not offered. FAVORABLE. Nothing. UNFAVORABLE. 1. Chlopin (p. 214) examined this color, and on his own experiments reports it as "strongly poisonous." On January 15, 1901, a dog weighing 18.4 kilos was given 3 grams. An hour or an hour and a half after giving dye vomiting Bet in; dog lay down on one side and died in 3 or 4 hours. Cause of death, heart filled with coagulated blood; lungs, liver, and kidneys filled with blood; turbid swelling of the liver and heart : mucous membrane "t" stomach inflamed ; l>rain unchanged. Death caused l.y paralysis of the heart. It also ads severely on the skin. G. T. 576. Tradr names. — New Gray; Malta Gray; Nigrosin; Direct Gray; Methylene Gray ; New Methylene Gray, Scientific name. — (?) Discovered. — 1888. s/i. "''' examined this color, and on his own experiments concludes it contains '"no poisonous properties " The experimental data follow : 136 COAL-TAR COLORS USUI) IN FOOD PRODUCTS. Experimental data ly Chlopin. [1 gram=77 ing=.">4 grains.] Date. Dose. Weight. **™* General condition of animal and urine. 1902. Feb. 8 9 Grants. Kilos. cr. 3 12.9 510 4«n Dm quite well; urine quite normal. Urine green color; acid; no albumen. Do. Do. Do. Vomited once; mine dark green; acid; no albumen. Do. Do. Urine lighter shade. 10 11 12 13 3 3 3 3 3 3 400 450 475 14 IS 16 "i.'.'.s 490' ]7 47.") Urine normal; dog well. Total . 21 G. T. 584. Trade names. — Safranin; Safranin S; Safranin cone; Safranin GOO; Safranin T; Safranin extra G; Safranin FF extra; Safranin AG, AGT, and OOF. Name under which it was offered on the United States market as a food color in 1907. — Safranin SP. Scientific name. — Mixture of diamhlo-phenvl-and-tolyMolazonium chlorids. Discovered and patented. — 1859. Shade. — Reddish brown. Offered bv 1 out of 12 sources. FAVORABLE. 1. Cazeneuve (Arch. gen. tie mid. 1886, Vol. I, disturbances but is not a violent poison. p. 753): Produces gastrointestinal ( \ FAVORABLE. 1. Prohibited by Confectioners' List. 2. Wf.yl (p. SI): "Caaeneuve and I .opine pointed out the poisonous nature of * * * Sairanine." 3. "This body (Chamber of Commerce of Soonebecg) recommends for the prepara- tion of children's toys tlir colors, the poisonous character of which 1 can demonstrate. These arc * * * Safranine * * * . (p. 34.) A. W'kvi, (Handbueh der Hygiene): "According to Theodore Weyl this is even in small doses, when injected Bubcutaneously, a strong poison" (60 milligrams per kilo body weight, or 35 grains per LOO pounds); "whereas, when administered by tlic stomach onlj large doses over a long period of time produce diarrhea." 5. Lbwim (Lehrbuch der T days, which amounts t<> 91 milligrams per kil<> per day, or 01 grains per LOO pounds per day. There was no disturbance o! any kind. Killed by chloroform; autopsy showed everything to be normal. G. T. 600. Tradi nanus. — Nigrosin, spirit soluble; Coupler's Blue; Oil Black; Sloelin; Spirit Black. Scientific name. — Mixtures of Iiuhilins with allied bases and fluorin- dins. I> icovered and patented. ls<'»7. Shade.— Black. Nol offered. i \\ ORABLK. I Permil '<•4 mg= 38 grains.] Date. Dose. Weight. 24 hours' urine. General condition of animal and urine. 1902. Jan. 5 6 Grams. 3 Kilos. 18.4 cc. 650 670 400 410 400 400 500 560 650 620 Dog and urine normal. Urine brown; acid; no albumen. 7 3 3 3 3 3 Urine greenish brown; no albumen. Do. g 9 Do. 10 Do. 11 Do. 12 Do. 13 Do. 14 Color and composition of urine normal; dog is well. Total . 18 No. 2. [1 gram=114 mg=80 grains.] 1902. Jan. 14 15 16 18 19 20 21 22 23 24 25 2(1 Total . a 3 3 8.8 420 390 420 3 3 3 8.8 400 460 400 450 4 18 Dog and urine normal. Urine slightly blue; acid; no albumen. Do. Do. Urine almost normal color; no albumen. Color and other properties of urine normal. No record. I 'line slightly bluish; no albumen. Do. Dark green; no albumen. Feeble green; no albumen. Urine normal; dog well. UNFAVORABLE. Santori (MolcschoWs Untertuehungen, 1895, v. 15, p. 55): A dog weighing 4,500 grams received 4 to (> grams dye in 7 days, which amounts to ]27 to 100 milli- grams per kil'> per day, or 89 to 133 grains per 100 pounds per day. No vomiting. Stool black hi nc and no change in the urine. Up l<> the sixth day the animal, in very good general condition, ate heartily and was quite lively; temperature Unchanged , On the morning of the sixth day the animal was found in his cage Buffering from general muscular cramps which were heightened by the slightest noise; the animal did not respond to calls or threatening movements and was in a complete stupor. This continued for 24 hours, when the animal died. Tho autopsy showed numerous punctures of I he lungB and of the mucous membrane of the stomach; Eatty degeneration of the liver and little blood in it; kidneys without change, although the COrtei wai colored a light green. COMPILED DATA UNDER GREEN TABLE NUMBERS. G. T. 602. 139 Trade names. — Nigrosin, soluble; Bengal Blue; Gray R and B. Scientific name. — Sodium salts of sulphonic acids of spirit nigro- sins. Discovered and patented. — 1867. Shade. — Black. Not offered. Nothing. FAVORABLE. UNFAVORABLE. 1. Lewin (Lehrbuch der To.rikologie, 1897, p. 931) Bays: "Produces eczema," and cites Deutsche. Med. Woehenschr., 1891, page 4~>- G. T. 614. Trade names. — Magdala Red; Naphthalene Rose; Sudan Red; Naphthalene Red; Naphthylamin Fink. Scientific name. — Mixture of amido-naphthyl-naphthazonium- chlorid and diamido-naphthyl-naphthazonium chlorid. Discovered. — 1868. Shade. — Red. Not offered. Nothing. FAVORABLE. UNFAVORABLE. 1. Chlofin (p. 100) examined this color, and on his own experiments concludes it "does not belong to the poisonous list, but is not wholly harmless." The ex- perimental data are as follow-; / zperimental data by Cfdopin. No. 1. [t gram— 91 mg— 64 grains.] D . Weight. 24 hours' urine. General condition of animal ;ml«.r: ;i« il; no line- Ill the morning thm< limes thin sIik»i, urine f.unii 1; no albumen. • ■ 7 i Do | Do 16 140 COAL-TAR COLORS USED IN FOOD PRODUCTS. Experimental data by Chlopin — Continued. [1 gram=114 mg=80 grains.] Date. Dose. Weight. 2 iirire^ i General condition of animal and urine. 1903. Feb. 10 Grams. 3 Kilos. cc. 310 315 350 n muriatic acid rose color, becomes greenish gray. 12 8.8 Xo vomiting; no diarrhea; color reaction of urine normal; no albumen. Total . 3 DoriiTIl I. 1. Wixogradow (Z/.s. Nahr. Genussm. l ( .)0.$, v. 6, p. 589) says it almost completely inhibits digestion. G. T. 620. Trade names. — Gallocyanin DII and BS; Fast violet; GallocyaniD RS, BS, and D. Scientific name. — Dimethylamido dioxy phenazoxoniumcarboxy- late. (BS is the bisulphite compound.) Discovered and patented. — 1881. SJiade. — Bluish Violet. Not offered. Nothing. KAVOKAHI.K. UNFAYOKAB1.E. 1. Sastori ( MoleschoW s Untersuchungen, 1895, v. 15, p. 51): A Jog weighing 5,400 grams received 7.5 grams dye in 30 days, which amounts to 4G milligrams per kilo per day or 32 grains per 100 pounds per day. Throughout all the time the animal remained well, had good appetite, temperature normal, no Loss of weight; urine and feces colored deep blue black. Killed by chloroform. Autopsy showed incipient fatty degeneration of the liver and a swelling of the kidneys. Conclusion: Poisonous. G. T. 639. Trade names.— Meldola's Blue; Cotton Blue Rj Fasl Navy Blue R; Naphthol Blue R and J); Naphthylene Blue R in crystals; Fast Blue K, 2 R and 3 R for cotton in crystals; Fast Navy Blue KM and MM. Scientific name. — Zinc double chlorid of dimethylamido-naphtho- phenoxazonium chlorid. Discovered . — 1 8 7 ( . I . Shade.— Dark Violet. Not offered. Nothing. I \\0(i grams received 12.5 grams dye during 30 days; this amounts to 93.3 milligram* per kilo per day or 65.3 grains per hundred pounds per day. Continued vom- iting beginning with 0.2 grams dye, anaemia and copious as well as continuous salivation and emaciation; the animal lost 1,200 grams in weight or 27 per cent. Killed by chloroform; autopsy showed no fat, flabby muscles, stomach contracted and filled with mucous and in part colored pea green; fatty degen- eration of the liver; contracted bladder; kidneys swollen and congested with blood and decomposed blood corpuscles in the Bowman capsules. G. T. 649. Tradenames. — Gentianin; Gentians Violet Scientific name. — Zinc double chlorids of nazthionium chlorid. Discovered he- Nothing. PAVORABl B. t \i A\ OKA HI I l. Santorj i MolfchotV* UnUrtuchungen, 1895, v. IS, p -. : \ dog w< ighing 3,008 grama received 4.7 grams of dye in 7 days, which amounts to 22 1 milligrams jht kilo per day or 157 grain- per LOO pounds per day. Beginning with the third day tii.' dog appeared weak and depr es s e d and ■ whitish troth appeal the mouth; mild diarrhea and complete aversion to food; temperature and urine unchanged. Died on seventh day. Autopsy showed congestion «/ mucous in. iic machj the liver was inflamed and the kidneys Btrongl] congested. 142 COAL-TAR COLORS USED IN FOOD PRODUCTS. G. T. 650. Trade names. — Methylene Blue B and BG; Methylene Blue BB in powder extra D; Methylene Blue BB in powder extra; Methylene Blue A extra. Names under which it was offered on the United States market as a food color in 1907. — Methylene Blue B; Methylene Blue. Scientific name. — Chlorid or zinc double-chloric! of tetramethyl- diamido-phenazthionium . Discovered. — 1876. Shade. — Blue. Offered by 2 out of 12 sources. 1AVORABLE. 1. Schacherl (p. 1046): "To these groups belong the much-used Methylene Blue, which in moderate doses is harmless." 2. Fraexkel (p. 574): "* * * Methylene Blue causes no noteworthy disturb- ances." 3. Cazexeuve (Arch. gen. de. med. 1886, v. 1, p. 753) says that it produces gastric intestinal derangements but is not a violent poison. unfavorable. 1. Prohibited by Confectioners' List. 2. Weyl (p. 31): "Cazeneuve and Lepine pointed out the poisonous nature of * * * Methylene Blue * * *." 3. Fraenkel (p. 579): "To regard Methylene Blue as a specific remedy BUch as quinin, is, in spite of a few such experiments, improper; it produces sub- sidiary effects which depend in part upon local irritation of the intestinal tract , and partly, however, upon irritation of the bladder with increased micturi- tion." 4. Cazexeuve (Arch. gen. de med. 1886, v. 1, p. 753), says that it produces gastric intestinal derangements. 5. Combemale and Francois (Sem. Med. 1890, no. 31, p. 258), say that it produces intestinal disorders and vomiting, colored urine, and colored feces in dogs, and therefore is a highly injurious color. 6. Saxtori (Molcschott's Untersuchunge?i. 1S95, r. 15, p. 4 J) classes it as injurious. A dog weighing 1,000 grams received in 20 days 18 grams dye, which amounts to 196 milligrams per kilo per day, or 137 grains per 100 pounds jht day. Urine and feces colored; diarrhea and continuous vomiting; blood and pus contained in stool; loss of appetite; loss of weight was 1 ,000 grams, or :r> per cent. Animal died. Autopsy disclosed blue-colored skin and kit; brain turned blue on exposure to air, but only the outer cortex was colored; Btomachical catarrh; the mucous membrane of the stomach colored blue; the heart sac and the pleura colored blue to blue-green; Intestines externally blue; kidneys were thickened and colored dark blue throughout; parenchymatous kidney inflam- mation; fatty degeneration of the liver; diaphragm Locally colored. 7. Aki.oivo and Cazbneuve, quoting Caceneuve and Lepine (Arch, de med. v. 9, p. 564 ', Bay thai it is not inactive. 8. Forbidden by Resolutions of Swiss Analytical Chemists, September, 1891. COMPILED DATA UNDER GREEN" TABLE NUMBERS. 143 9. Galliard (Rev. intern, des falsifications; abst. Hygien. Runsdach. 1892, p. 104)'- " Methylene Blue, which is frequently used for coloring foods, can cause, even in small doses (10 to 20 milligrams, or one-sixth to one-third of a grain), a feel- ing of general depression, nausea, and certain feelings of pain, and can even produce transitory albuminuria. In larger doses (40 to 60 milligrams, or mx- tenths to nine-tenths of a grain), it causes in the case of persons not accustomed thereto, vomiting, diarrhea, increased micturition, and albuminuria. In the case of persona suffering from nervous diseases, it frequently produces disturb- ance which has as its consequence a cessation or a change of place of the pain Sometimes it produces pain] ing of pain in the patient, for which no certain therapeutic indication could be determined." 10. Weyl (Ila/tdbuch drr Hygiene) comments as follows "These statements of Gal- liard arouse but little confidence, because, as is well known. Methylene Blue Lb very frequently administered to invalids in doses of more than 0.5 gram without any noticeable disturbance. Perhaps Galliard 'a preparation was unclean. At any rate, in all experiments on the poisonous nature of Methylene Blue it is to be considered that it frequently occura in commerce as a zinc chlorid double salt." 11. Lewin {Lehrbuch dcr Toiilologie, 1897, p. 232): "Methylene Blue * * * can produce after prolonged administration of 0.5-1.5 gram daily increased micturi- tion, irritation of the bladder, blue coloration of the urine, and .-aliva, diarrhea, headaches, vertigo, delirium, and twitching of the muscles, the latter symp- toms probably because the coloring matter is deposited in the brain." 12. Buss tiste it as poisonous. G. T. 651. Trade name. — Methylene Green G cone, extra yellow shade. Scientific name. — Xitromethvlene Blue. Discovered and patented, — 1SS6. Shade. — Green. Not offered. Nothing. FAYORAHI.F. • VOHAHl.K. 1. (hi. <>ri\ pp examined this color, and on his own experiments concludei ii is "non poisonous, but somewhat suspicious.' 1 The experimental data are as toll • inu ntal data by < 'hlopin. Nn. I. ins.) ' '. ^JSe?' Imaland urine. 1901. Oram*. h ii h - li: urine m i rlne dai k . no albumen. bumen. i rlne dark brown; no M <8fi l - ; . : n< ■ Total . •* 144 COAL-TAR COLORS USED IN FOOD PRODUCTS. Experimental data by Chlopin — Continued. No. 2. [1 gram=122 mg=X5 grains.] Date. Dose. Grains. 3 ! 3 3 Weight. urine* General condition of animal and urine. 1901. LVov. 25 26 Kilos. 8.2 cc. 430 445 450 410 440 Dog well; urine normal. Urine strongly green: no albumen. Do. 27 :::::::::: 28 29 Diarrhea; urine green: acid; no albumen. Hec. 12 Color normal: acid: no albumen: dog lively and well. Total. 15 DoriiTI-'l I.. 1. Win'ogradow (Zts. Nahr. Genussm., 1903, v. 6, />, It noticeably retards digestive action; is not indifferent . G. T. 654. Trade name. — Toluidin Blue O. Scientific name. — Zinc-double-chlorid of dimethvl-diaiuido-tolu- phenazthonium-chlorid. Discovered and- patented. — 1888. Shade. — Blue. Not offered. Nothing:. FAVORABLE. UNFAVORABLE. I. Fraenkel (p. 374)'. "Toluidin Blue * * * is a strong poison for micro- organisms, and may be used in eye treatment, like Methylene Blue." G. T. 659. Tradi names. — Primulin; Thiochromogen; Sulphin; Polychromin; Aurcolin. Scientific name. — Sodium salt of the mono-sulphonic acids of the dehydrothionated condensation products of dehydrothiotoluidin (mixed with some sodium dehydrothiotolwdin-eulphonate). Discovered. —1887. Shade. — Yellow. Not offered. Nothing. I W ORABLE I MA\ oi; M.I I I Chlopin (p. 103) ex/unined this color, and on his own experiments reports it as 'rapicious." The experimental data are as follows: COMPILED DATA UN DEB GREEN TABLE NUMBERS. Experimental data by Chlopin. [I gram=154 mg=l()S grain.-.] 145 Date. Dose. Weight. 24 hours" urine. (Sonera! condition of animal and urine. 1902. Mar. 21 22 23 Grams. 2.0 id Kilos. 0.5 cc. 292 294 355 200 Dog and urine normal. Urine dark brown: insignificant traces of albumen; acid. 24 25 i 0. o 0.3 L'rine dark brown; acid; no albumen. Do. 20 No albumen: dark-brown color: dog has lassitude; does not eat bread nor take milk, only a little meat. No albumen; dark-brown 'color; dog has lassitude; dotj eats little. At the point of injection an abscess appears; urine . albumen; yellow color. Urine acid; no albumen; yellow color: dot; eats more. 27 291 28 29 270 280 30 Apr. 2 Urine normal: appetite almost normal. Total . 1 '5.0 { .0., 1 liv mouth. - Subcutaneously. al*o Chlopis (Zts. NaJir. Genuasm., t902, v. 5, p. ?41). I><) l IJTFt 1 I. Winogradow Zts. Nahr, Genussm., 19Q3, v. 6*, p. 589) says ii noticeably retards digestive action; lw<>\ indifferent. G. T. 667. Trad* names. -Quinolin Yellow; Quinolin Yellow, water-soluble. Narru under which it was offered on the United States market as a food color t/i 1907. — Ckinolin Yellow O. Scientifr name. — Sodium salt of the Slllphonic acid (chiefly disulphonic acid) of quinophthalone. Discovered. 1882. Shade. — Greenish Yellow. (Ml'crcd by 1 out of 12 source Nothing. I'AVOKA BL1 i\ I A \ ()lt.\ BL1 I Chlopin (p. W5 1 on his own experiments reports this color as "suspicious." Tho experimental data arc as follow.-: ital data by t 'Mopin . No. l. [i gram ii tng 81 gralna | Jl hours' iinm-. '..-n.-ial condition «.f :m;in:il :md HI 790 • experlmenl dog and urine normal i rinc .in k brown; traces albumen >l uriiK- normal; do albumen. •J l 97291°— Hull. 147 -12 10 146 COAL-TAR COLORS USED IN FOOD PRODUCTS. Experimental data by r hlopin — Continued. Weight. 24 hours' urine. General condition of animal and urine. Mar. 23 Grams. 0.75 Kilos. 21.4 cc. 750 Urine dark brown: no albumen. Dog eats nothing. Do. - 603 725 5 Do. 27 Dog eats nothing: eats poorly; albumen in urine. Dog eats nothing; rather much albumen. Do. 30 Do. Apr. 2 Dog eats nothing; traces of albumen. i 3.00 710 Color of urine almost normal; insignificant traces of albumen. Urine yellow brown; traces of albumen; appetite normal. Trine Yellow brown: no albumen. - 700 Urine and dog normal. Total. | i 6. 00 \ 2 0. 75 No. 2. [1 gram= 133 mg=*83 grains.] 1903. Mar. 3 4 5 6 3.00 3.00 3.00 2.00 , Total.. 11.00 320 Before experiment dog quite well, and urine normal. 300 Color dark yellow; no albumen; general condition normal. 300 I Color yellower than normal; acid; no albumen. Do. 1 By mouth. 2 Subcutaneously. DOUBTFUL. 1. Wznooradow (Zts. Nahr. Genussm., 1903, v. 6, p. 589) says it noticeably retards digestive action; La not indifferent. G. T. 670. Trade names. — Vidal Black; Vidal Black S. Scientific name. — Possibly the sulpho-hydro derivative of a polythiazin. Discovered. — 1893. Shade. — Green. Not offered. Notl 1. Ch i \ VORABLB. i \i \ VOB \ hi i . iiii'is i />. 108) examined this color, and on his own experiments classes it as ■ very harmful. " The experimental data are as follows. Experimental data by Chlopin. [i gram - 132 mg 92 grains.] -' 3 _m hours' urine. 7.e 890 General condition of animal end urine. i m.i 1 1 dog quits normal; toon altar taking dys vomited tu Ice. Qeneral oondltion good; no vomiting; urine turbid, alkaline; no albumen. COMPILED DATA UNDER GREEX TABLE NUMBERS. Experimental data by Chlopin — Continued. 147 Date. Dose. Weight. urine"* General condition of animal and urine. 1901. Dec. 31 Grams. 3 3 Kilos. cc. Vomited soon after getting dye; appetite less: no albumen. Urine turbid; scarcely noticeable blackisb sheen; green with sul- 1902. Tan. 1 2 7 358 370 phuric and hydrochloric acid. Do. 4 3 3 Vomited all day after taking dye; eats little: no albumen. 5 Violent and prolonged vomiting and diarrhea; condition piti- able; further giving of dye stopped, so as not to kill the animal. Vomiting and diarrhea continue. 5-7 8 Vomiting and diarrhea stopped. Dog is livelier, and begin to eat. Dog looks well; urine normal in color; acid; no albumen. 9-11 13 Total.. 17 G, T. 675. Trade names. — Thiocatechin; Thiocatechin S. Shade. -Brown. Xot offered. Nothii FAVORABLE UNFAVORABLE 1. Chlopin (pp. 'in. til examined this color and classes it as '"very poisonous. Tl •■ experimental data arc as follows: Experimental data by Chlopin. [i gram 12s nig— 90 grains.] General condition of animal and urine. Dene. Weight. 24 hours' urine. 1902. Feb. 13 14 Grams. 2.4 2.0 (I.:, Kilo*. 7.S cc. 300 300 u LI is 4.9 Before experiment dog and urine quite normal; 10 or 15 minutes after giving dye dog fell on one side, limbs extended, stomach drawn in; small and frequent convulsions: retching; then abundant vomiting, same color as dye; soon after vomiting dog got up and walked as if drunk; hind legs tend to collaj. ration. a few minutes after giving dye, dog again fell as if in an epileptic fit; convulsions of the extremities, which » but the dog si ill lay stretched out, with open eyes, which reacted to light: tongue banging out to one side; after 10 or US minutes vomiting began; dog Mill lying on one Bide, assumed a more normal attitude: 15 minutes later dog got up. walked as if drunk; poor control over hind Vomits at once after getting dye, but remained standing. Dog appears dep re s se d, bul Dog quite well; urine usual normal color: no albumen. Tradi names, Sci< itt'ific nana Shade. Blue. G. T. 689. -Indigo; Indigo Pure BASF. — [ndigotin. \<»t offered. 148 COAL-TAR COLORS USED IN FOOD PRODUCTS. FAVORABLE. 1. Permitted by Confectioners' List . 2. Permitted by the law of Italy. UNFAVORABLE. 1. Fkaknkfl (p. oil): "However, pure Indigo, according to Robert, is, in finely divided condition, a violent local irritant." G. T. 692. Trade names. — Indigo Carmine; Indigo Extract; Indigotin. Xamrs under which it was offered on the United States market as a ft>od color in 1907. — Indigo Carmine Powder IN; Indigotin; Indi- gol ni A. Scientific name. — Sodium salt of Indigotin disulphonic acid or the in-o acid. Discovered, — 1740. Shade. — Blue. Offered by 3 out of 12 sources. FAVORABLE. 1. Schacherl (p. 1046): " No objection to its use." 2. Santori ( .Volcschott's Untcrsuchunyen. 1895, v. 15, p. 41): A dog weighing 4,500 grams received 90 grams dye in 30 days; vomited twice during the examina- tion; no change in weight; animal killed with chloroform; autopsy showed -light dull swelling in the epithelium and convoluted canals of the kidneys. This dosage amounts to 6(!7 milligrams per kilo per day, or 4G7 grains per 100 pounds per day. Classes it as harmless. ALPHABETICAL INDKX OF TRADE NAMES OF COAL-TAB COLORS. The following list of the trade names of coal-tar colors appearing in the foregoing compilation on physiological action is complete when supplemented by the list of 23 colors given on page 227; Green Table numbers in parenthesis. A- (tin Blue (599). Acid Brown (138). Arid Green I 134,435). Acid Green ex. cone. (435). Acid ( rreen cone. I 135). Acid Green cone. \'\ (435). Acid Green cone. 790 | 135). Acid Fuchsin ( 162). Arid Magenta I 162). Acid Magenta powd. (462). Acid Roeein | 162 1. Acid Rubin I 162). Acid Violel I W Acid Violet 17). Eoan GG i 512). 150 COAL-TAR COLORS USED IN FOOD PRODUCTS. Eosin J (512, 517). Eojrin 10 B (521). EoeiD 3J & 4J ext. (512). Eocrin Y (512). Eo^in Yellowish (512). Erika B (78). Erythrobenzin (448). Erythrosin (517). Erythrosin B (517). Erythrosin D (517). Erythrosin G (516). Erythrosin yellow shade (517). Ethyl Green (428). Fast Blue (477). Fasl Blue B, spirit soluble (599). Fast Blue R and 3R (601). Fast Blue R, spirit soluble (599). Fast Blue2R, B, and 6B (60! Fast Blue R, 2R, and 3R (639V Brown G (138). Fast Green (427). Fast Green J (428). East Green O (394). East Myrtle Green (394). Fast Navy I'.lue R (639). East Navy Blue RM and MM (639). Ponceau B | L6 East Red (102, 105). Fasl Red C (103). Fast Red D (107). Fasl Red EB (107). Fast Violet (620). Fellow (4,8, 9,88). Fasl Yellow extra (8). Yellow G Fast Yellow 053 (8;. Yellow It (9). Fine Blue (457). Fuchsiacin (448). Fuchsin I I Fuchsin cryst. (448 Fuchsin 8 (462). Gallocyanin I'll and us (620). Gallocyanin lis. BS, and D I Gentian Bluet tiana Viol< I ttianin (64 Gold Orange - Gold Yellow (84 Golden Yellon Rand B I Green E G sen 087 12* Guin< Harmalin (448). ilelianthin (87). Helvetia Blue (479). Hessian Blue (457). Hofmann Violet (450). Hydrazin Yellow (94). Imperial Scarlet (163). Imperial Yellow (6). Indigen D and F (599). Indigo (689). Indigo Carmine (692). Indigo Carmine powd. IN (692). Indigo extract (692). Indigo pure BASF (689). Indigotin (692). Indigotin A (692). Indisin (593). lndisinRandB(601). Indophenol (572). Endophenol white (572 . Indulin opal. (599). Indulin sol. (601). Indulin, spirit soluble (599). Indulin 3B opal. (599). Indulin 6B opal. (599). Indulin 3B%p. sol. (599). Indulin 6B sp. sol. (599). Indulin 3B (601). Indulin 6B (601). Iodeosin B (517). [odeosinG (516). Eodin Green (459). lodin Violet (450). Jaime Acide (4, 8). Jaune Acide C (-1 >. Jaime Xaphthol (3). Jaune d'Or Jaune d'Orient (516). Jaune Soleil (39 Kaiser Yellow Kermesin Orange (97). Leather Brown (197). Leather Veil-, Lemon Yellow (4). Light Green 8 F bluish (434). a s F yellow shade | 135), London Blur extra Magdala Red (614). Magenta < 148 M , pnta I' A B 8 Red (448). ,ia Powder A (448 Maize (399). Malachite Green (42' Malachite Green B (427-428). COMPILED DATA UNDER GREEN TABLE NUMBERS. 151 Malta Gray (576). Manchester Brown (197). Manchester Brown EE (201). Manchester Yellow (3). Mandarin G R (97). Mandarin G ext. (86). Martius Yellow (3). Mauve (593). Mauve Dye (593). Mauvein (593). Meldola's Blue (639). Metanil Yellow (95). Methyl Blue water soluble (478). Methyl Blue for silk MLB (478). Methyl Blue O (479). Methyl Violet (451). Methyl Violet B (451). Methyl Violet BB ext. (451). Methyl Violet 3 BD (451). Methylene Blue (650). Methylene Blue A ext. (650). Methylene Blue B and BG (650). Methylene Blue B D (650). Methylene Blue BB extra (650). Methylene Blue BB extra D (650). Methylene Gray (576). Methylene Green G. cone. ext. (651). .Naphthalene Pink (614). Naphthalene Red (614). Naphthalene Rose (614). Naphthol Black B (188). Naphthol Black BDF (188). Naphthol Blue R&D (639). Naphthol Green (398). Naphthol Green B (398). Naphthol Grange (85). Naphthol Red S (107). Naphthol Yellow (3, 4). Naphthol Yellow L (4). Naphthol Yellow 8 (4,5). Naphthol Yell.,wSLOZ(4,86). Naphthylamin Yellow (3). Naphthylene Yellow (2). Navy Blue B New < loccin (101 New Gh New Green I 12 New If ethylene Gray (570). New Red L | Lfl \ ictaria Green r_':.428). New Yellow (88). New Yellow L - Nicholson Blue 177). Nigh! Green (460). Nigrosin (576). Nigrosin sol. (602). Nigrosin sp. sol. (600). Nitrodiphenylamin (6). Nitromethylene Blue (651). Oil Black (600). Oil Orange 7078 (11). Oil Yellow (16). Opal Blue (457). Orange (86). Orange I (85). Orange II (86). Orange III (87). Orange IV (88). Orange A (86). Orange A extra (86). Orange Brown (17, 18, 41). Orange A 1201 (86). Orange B (85). Orange extra (86). Orange G (14). Orange GG (14). Orange GG Crystals (14). Orange GRX (13). Orange GS (88). Orange GT (43). Orange M (88). Orange MN (95). Orange N (43,88). Orange O 27 (85). ge O (43). Orange R (15, 55, 97). Orange RN (43). Orange RZ (85). Orange T (97). Orange 2 R (97). Orange Y (86). Orcellin No. 4 (102). Paris Violet (451). Phenylene Brown (197). Philadelphia Yellow G (532). PhloxiinVj] . Thloxin T Phoepbin 5 Picric Add Polychromin (660). Pomona Gn Ponceau B (163). Ponceau A GB (IS). Ponceau 4 KB mi (. and ( . R Ponceau R ; ma kb 152 COAL-TAR COLORS USED IN FOOD PRODUCTS. Ponceau 2 G (15). Ponceau 2 R (55;. Ponceau 6 RB (109,. Primula (450 . Primulin (6« Printing Blue (599;. Pure Soluble Blue (480). Pyoctanin Aureum (425;. Pyoctanin (451). Pj rosin B (517). i\ r -in J (516 . Quinolin Yellow, water-soluble (667). Quinolin Yellow (667). Rauracienne (102). Red (107). Red Violet 5 R extra (450). Resorcin Yellow (84). .■in 275 (84). Rhodamin (504). Rhodamin B (504). Rhodamin O (504;. Rhodamin B extra (504). Rhodamin G and (i extra (502). Roccellin (102,. Rose Bengal (520). Etoee Bengal AT (520). Rose Bengal G (520). Rose Bengal N (520). I' •••in (448;. Rosolic acid (483). Rubianite (448). Rubidin (102 . Rubin (448). Rubin S (402). Russian Green (394). Saffron Substitute (2). Saffron Yellow (3, 4). Bafranilin (504). Safranin (584). Safranin AG, A.GT, and OOF. (584). Safraniii Cono. (584). Safranin extra < I 1 58 1 !.in PF extra (584;. Safranin GOO. < Safranin S. (584). Safranin SP Safranin T 584 Scarlet (55). Scarlet L. (106). Sloelin (600). Sloelio RS. and BS. (001). Solferino (448 Solid bellow i i Soluble Blue | L£ Spirit Black (600). Spirit Blue (457). Soluble Blue 8 B. (479). Soluble Blue 10 B. (479). Soluble Blue XL. (479;. Succinic (4). Sudan I (11). Sudan Red (614). Sulphin (659). Sulphin Yellow (4). Sulphonaphthol Acid Yellow (4). Sultan Red 4 B. (277). Sun Yellow (399). Tartrazin (94). Thio Catechin (675). Thio Catechin S. (675). Toluidin Blue O. (654). Tropaeolin D. (87). TropaeolinG. (95). Tropaeolin O (84). Tropceolin OO (88). Tropseolin OOO (85). Tropaeolin OOO No. 2 (86). Tropa?olin R (84). tfreol D. (574). Ursol DD. (574). Ureol P. (574). Vert Diamant (427). Vert Lumiere (459). Vesuvin (197). Vesuvin B. (201). Victoria Blue B. (488). Victoria Blue 4 R. (490). Victoria Orange (2). Victoria Yellow (2). Victoria Yellow Cone. Z. (95). Vidal Black (670). Vidal Black S. (670). Violein (593). Violet de Methylanilin (451). Violet R. (450). Violet RR. (450). Violet 5 R. (450). Water Black (166;. Water Blue 6 B extra (ISO). Water Blue (180). Xanthin (532). XL Soluble Blue (479). Xylidin Red (55). Xylidin Scarlet (55). Yellow Corallin (483). Yellow FY. (4). Yellow MXX Cone. (95), Yellow W Yellow Wit. (89). DOSAGE AND SYMPTOMS. 153 X. DOSAGE AND SYMPTOMS. CONFECTIONERS' LIST AS A BASIS FOR A RULE. Considering the Confectioners' List of 1899 as a correct guide as to which colors are harmful and which are harmless, the attempt has been made to determine how far dosage and the corresponding physiological effects may serve as a guide in determining which colors, other than those enumerated in either portion of the Confectioners' List, are harmful or harmless. (See p. 48.) To this end the available literature has been searched and classified, and wherever it was possible to arrive at any conclusion as to the actual dose or the average dose over a stated period of time, and the corresponding physiological observations, these data have been sepa- rated and brought together for the purpose of making comparisons and deductions therefrom. It w r as thought that the literature would show that if a dog or other animal is killed by a certain given amount of color per 100 pounds of body weight of the animal that such color is always harmful; that if untow T ard effects, such as vomiting, diarrhea, weakness, and general depression, are caused by more than a certain weight of color per 100 pounds body weight of the animal, such color is always regarded as harmless. The classification of the available literature and the conclusions therefrom are as follows: Of the 33 coal-tar colors listed as harmless in the Confectioners 1 List, 10, namely, G. T. 4, 9, 55, 65, 85, 103, 105, 107, 448, and 462, have been tested on humans, while the conclusion as to the remaining 23 is reached by the effects observed on dogs alone. However, con- tradictory statements are recorded in the case of No. 9, and none of these tests was of long-continued duration, but, on the contrary, in many cases the time covered was exceedingly short, and the conclu- sions deduced are, therefore, not necessarily final nor correct. It should he further noted that No-. 95 and 106, reported as nonpoi- sonous to humans, are in the harmful section of this Confectioners' List . I. Those colors which produced no effect are as follows (the num- ber of grains given i- the amount administered per 100 pounds body weight : where the data permitted, the number of days' duration of I he experiment is also given : . Brilliant Fellow 8 13. Ponceau 4 GB LIS (21 TlilMxin 521. Phloxin BOO 154 COAL-TAR COLORS USED IN FOOD PRODUCTS. II. Those colors which in some cases produced no effect, and in others produced effects, are as follows: 4. Naphthol Yellow 395 grains produced diarrhea and no albuminuria. 292 grains produced albuminuria. 116 grains for 25 days produced no effect. 25 grains for 2 weeks, on alternate days, produced no effect. 8 and 9. Fast Yellow Y and R: 173 grains for 3 weeks produced no effect. 53 grains produced albuminuria. 55. Ponceau 2 R. 582 grains killed. 198 grains produced no effect. 05. Fast Red B: 143 grains for 145 days produced no effect. 137 grains for 145 days produced no effect. 98 grains for 145 days produced no effect. 68 grains produced albuminuria. 38 grains produced albuminuria. 103. Azorubin S: 143 grains for 145 days produced no effect. 137 grains for 145 days produced no effect. 98 grains for 145 days produced no effect. 70 grains produced diarrhea. 287. AzoBlue: 233 grains produced no effect. 166 grains produced albuminuria. III. Those that produced only a slight disturbance which was regarded as negligible: 28. Archil Substitute V: 301 grains produced vomiting and albuminuria. 127 grains produced vomiting and albuminuria. 105. Fast Red E: 70 grains produced diarrhea. 166. Wool Black: 117 grains produced albuminuria. 240. Congo Red: 192 grains produced albuminuria. 269. Chryaamirj R: 433 grains produced albuminuria. 361 grains produced albuminuria. 39 1. Dinitrosoresorciii : L39 grains produced loss of appel Ite. 121 grains produced albuminuria. From Class 1 above it would appear — (a) Thai if a dose of 1 13 grains per 100 pounds body weight pro- duced no effect it is to be classed as harmless. From Class II above Li would appear — (a) That if albuminuria is caused by as little as 38 grains in some cases, and no effect is produced by doses as large as 14.'* grains, the volor is to be classed as harmless. DOSAGE AND SYMPTOMS. 155 (6) Also that if 198 grains produce no effect the color is to be classed as harmless, even if 582 grains in another case kill the animal. (c) That if small amounts produce albuminuria and larger amounts do not the color is to be classed as harmless. (d) That if small amounts produce diarrhea and larger amounts do not the color is to be classed as harmless. From (lass III it would appear that — (a) Albuminuria produced by as little as 1 1 7 grains is to be regarded as not harmful. (b) Diarrhea produced by 70 grains is to be regarded as not harmful. (c) Vomiting and albuminuria produced by 127 grains are to be regarded as not harmful. The colors classed as harmful in the Confectioners' List may be classified as follows : I. Dogs. — Those colors that under certain conditions produce no noticeable effects on dogs, while under other conditions effects are observed, are classed as harmful colors: 1. Picric Acid: Dogs stand 5 grains daily for 18 days without effect. 18£ grains produced diarrhea. 20.4 grains killed. 2. Dinitrocresol: 140 grains caused vomiting but did not kill. 38£ grains caused vomiting but did not kill. 38£ grains caused vomiting and did kill. 35 grains caused vomiting but did not kill. 31.6 grains caused vomiting but did not kill. 3. Martins Yellow: 51 grains do not kill, but produce weakness, diarrhea, and albuminuria. 50 grains kill. 8 grains do not kill, but produce weakness, diarrhea, and albuminuria. B6. Orange II: 71 1 trains produced kidney irritation, thirst, and diarrhea. 333 grains kill, and produced diarrhea and albuminuria. 244 grains produced kidney irritation, thirst, and diarrhea. 36 grains produced DO effect. a"). Itetanil Yellow: 620 grains pr<>dnre vomiting. 803 grains kill. m? grains no effect , 371 grains kill. lot grains produced albuminuria. tiL* grains no effect •i I grains do effect , ::i grains kill. i!»7. Bismarck Brown: 2 16 grains produced vomit ing. I is grains produced vomiting and general depression. grains daily for 3o days, do effect. 156 COAL-TAB COLORS USED IN FOOD PRODUCTS. 398. Naphthol Green B: 600 grains produced green urine and conjunctiva. 292 grains produced no effect. 12S grains produced green urine and conjunctiva. II. Those colors that have not been observed to give negative results, but which have caused certain effects: 11. Sudan I: 118 grains produced colored urine, vomiting, and albuminuria. 17 and 18. Chrysoidin Y and R: 79 grains produced albuminuria. 74 grains for 30 days produced no albuminuria, but a loss of 12 J per cent of body weight. 88. Diphenylamin Orange: 216 grains produced albuminuria. 128 grains produced phenol in urine and albuminuria. 138. Fast Brown G: 237 grains produced diarrhea and loss of appetite. 218 grains produced diarrhea. III. Rabbits: 1. Picric Acid: 24.5 grains for 90 days, no effect. 45 grains for 19 days kill. 86. Orange II: 292 grains, no effect. 933 grains kill. 95. Metanil Yellow: 216 grains, no effect. IV. Humans: 1. Picric Acid: 8.3 grains no effect. 13.8 grains no effect. Invalids and children can not stand this color. 2. Dinitrocresol: 42 grains kill. 86. Orange II: 1£ grain.-, no effect. 3 grains, headache, vertigo, dryness of throat, and poor general condition. 95. Metanil Yellow: 1£ grains, no effect. 3 grains, n<> effect. From Classes I and II above, it would seem to appear — (a) That if a dog is killed the Colof is harmful, even though it take us much M 008 .-Mm-, or aa little as 20.4 grains per LOO pounds body weight t<» kill. (b) That if albuminuria is produced in dogs by as little as 7!) grains. <>r as nnicli M tins, the color is harmful. (c) That if diarrhea la produced in dogs by as much aa 218 grains, or is little m ">o as, ih<- color is harmful. (d) That even If a dog can take as much a.s 107 grains, the color may be harmful. DOSAGE AND SYMPTOMS. 157 From Class III it would appear — (a) That even if rabbits can stand as much us 292 grains, or as little as 24.5 grains, the color is harmful. From Class IV it would appear — (a) That if 42 grains kill a human, the color is harmful. (6) That even if humans can stand as much as 3 grains without untoward effect, the color is harmful. The conclusions that may be drawn from these data are: 1. If a dog is killed by — (a) 603 grains per 100 pounds body weight, the color may be harmful. (6) 582 grains per 100 pounds body weight, the color may be harmless. 2. If a dog can bear without effect — (a) 407 grains per 100 pounds body weight, the color may be harmful. (b) 198 grains per 100 pounds body weight, the color may be harmless. 3. If albuminuria is produced in a dog by — (a) 38 grains, or 143 grains per 100 pounds, body weight the color may be harm- less. (6) 79 grains, or 128 grains per 100 pounds body weight, the color may be harm- ful. 4. If diarrhea is produced in a dog by — (a) 70 grains per 100 pounds body weight, the color may be harmless. (6) 50 grains per 100 pounds body weight, the color may be harmful. 5. If vomiting and albuminuria are produced in a dog by 127 grains per 100 pounds body weight, the color is not necessarily harmful. 6. If small amounts of color produce in a dog diarrhea or albuminuria, and larger amounts do not, the color may be harmless. 7. Even though rabbits can withstand 292 grains per 100 pounds body weight, the color is not necessarily harmless, but may be harmful. 8. If 42 grains kill a human, the color is harmful. 9. If a human can withstand 3 grains without effect the color is not necessarily harmless, but may be harmful. 10. If a human can not withstand 3 grains, even though it can withstand 1^ grains, the color is not necessarily harmless, but may be harmful. Lehmann I Meihoden der praktischen Hygiene, 1890, p. 545) says: I regard such substances as harmful to health which when fed t<> a sound dog in doses of a few decigram^ per day produce at once, or after a few repetitioni of the dose, disturbance in the health of the dog; on the other hand, dyes which in doses of bom one to several grams can be taken for weeks '>n end withoill Causing any disturbance or only sli;_rlit intestinal disturbances "r a slight and passing albuminuria can bo regarded a- harmless, h. should never be forgotten that a few milligrams of a coal- tar color dyes very strongl) and it a not easy even by most extraordinary use of colored objects, e. g., by children that more than milligrams, at most centigrams, of dye can he introduced into the human Btomach. In extremely large doses many substances, for example, all our condiments, are naturally harmful. Of 65 dogs weighed ami experimented on by Chlopin the average weigh! waa , .».•- , kilograms, or 20.3 pound.-: almost half the d weighed between <"> ami 11 kilos. To adapt the rules of Leliiiinnu to a basis of grains per 100 poundsof body weight, assuming the average weight of a dog to be 20 pounds, 158 COAL-TAR COLORS USED IN FOOD PRODUCTS. it is only necessary to multiply the weight of color by 5 and by 15.432. Doing so, the following data are obtained: A. Harmful colors produce bad e fects in doses of a "few" decigrams, i. e., mul- tiples of a half gram; the word "few" is not sharply denned, but the rule means a "few" times 7.7 grains. B. Harmless colors produce, when continuously fed, no bad effect in multiples of 10 of the doses of A, above, i. e., multiples of 77 grains. I Slight diarrhea and slight, temporary albuminuria are not to count against the color. Applying these rules to the preceding colors discloses that they were apparently not so used, in selecting the harmless and harmful colors in the Confectioners' List. LEHMANN'S RULES. The Lehmann rules as just interpreted were applied to the data just given. As a result of such application the conclusion is reached that of the 15 colors classed as harmless (foregoing classes I, II, and III), 6 would be classed as doubtful by the Lehmann rules, namely: 13. Ponceau 4 GB, 65. Fast Red B, 8 and 9. Fast Yellow Y and R, 103. Azorubin S, 55. Ponceau 2R, 105. Fast Red E, and the remaining 9 would have been classed as harmless. Of the 11 classed as harmful (foregoing classes I, II, III, and IV), 2 would have been regarded as harmless, namely: 308. Xaphthol Green B, 138. Fast Brown (1, 3 would have been classed as doubtful, namely: 197. Bismarck Brown, 88. Diphenylamin Orange, 11. Sudan I, and the remaining 6 would have been classed as harmful. SANTORI'S WORK AS A GUIDE TO A RULE. Santori regards the following six dyes as harmless: 157 (18S), 467 (74), 477 (129.5), 4S0 (156), 599 (64), 092 (467). The first is the Green Table number and the bracketed figure the average Dumber of grains per 100 pounds per day for 30 days. In the case of Nbs. 457 and 599 the animal was wholly normal throughout the test and the autopsy showed only normal conditions. In the case of \o. 477 (he only disturbance w is colored feces and the autopsy showed only normal conditions. In tin 1 case of No. ISO the only disturbance was colored feces, but. the autopsy showed a pea-green kidney. In the case of Nbs. 167 and 692 there was only vomiting; in the of N<>. 167 the autopsy showed everything normal; whereas in the of No. 692 the autopsy disclosed a swollen kidney. Therefore, according to Santori, a dyv that causes colored feces, even with colored kidney, is harmless. Also a dye that causes vomit- OIL-SOLUBLE OK TAT COLORS. 159 ing and swollen kidney is harmless. On the other hand, a dye (572) which produces no change or symptom observable during life, but fatty degeneration of the liver is shown, at the autopsy, is a harmful dye. This lack of conformity makes the relation between dosage, symptoms, and harmfulness or harmlessness more confusing and per- plexing. YOUNG'S RULE. It must be remembered that smaller amounts of drugs, and, there- fore, of coal tar colors, effect children as a rule than are effective upon adults. Taking Young's rule as a guide, it appears that gen- erally the effective dose for a 3-year-old child is one-fifth the effec- tive adult dose; for a 4-year-old child one-fourth; for a 6-year-old child one-third; for an 8-year-old child two-fifths and for a 12-year- old child one-half the effective adult dose. All of this should be taken into account in drawing conclusions from experiments as to the harm- lessness of any coal-tar dye upon humans. Very little attention has, however, been paid to this aspect of the matter in spite of the fact that colored foods, confectionery, pastry, beverages, and the like are partaken of by children and in many cases such articles are prepared for the sole or particular consumption of the very young. From the foregoing data it seems clear that deductions as to the harmlessness or harmfulness of coal-tar dyes when administered to dogs, and not based upon autopsies, are not final nor conclusive as to the effect upon the dog. The extent to which such deductions are correctly transferable to humans is likewise not established. XI. OIL-SOLUBLE OR FAT COLORS. Oil-soluble colors are used for coloring fats, such as butter, oleo- margarine, edible oils, and the like; of the coal-tar colors the oil- soluble colors are chemically nonsulphonated azo-colors. The nonsulphonated azo-colors which have been physiologically examined arc: 11. Sudan 1 (anilinaxo-b-naphthol) ('J*. Hi. Butter Yellow (anilin-aso-dimethylanilin). 17. Chrysoidin V (anilin-aso-m -phen} lene-diamin 18. ChryBoidin R (aniliiia&o-m-tolylene-diamin) (1). 11. ChryBoidio R (o-toluidin-aso-m-tolylene-diamin). li)7. Bismarck Brown (m-phenylene^iajum-disftzo-m-phenylene^iamix) I 201. Manchester Brown (m tolj lene-oUamin^isaxo-m-tolylene-diamin The numbers preceding the trade Dames arc the Green Table num- bers; ili' 1 scientific aames appear in parentheses; and the number of dealers offering the colors on the United States market in the summer of 1907, out of a possible 12, appear after the scientific name, also in parentheses. 160 COAL-TAR COLORS USED IX FOOD PRODUCTS. For all but No. 201 contradictory statements occur in the compiled literature; for Xo. 201 only unfavorable reports were found. There were on the United States market in the summer of 1907 the following oil-soluble colors: 10. Sudan G (anilin-azo-resorcin) (1). 11. Sudan I (anilin-azo-b-naphthol) (2). 4U. Sudan II (xylidin-azo-b-naphthol) (1). 60. Carminaph Garnet (a-naphthylamin-azo-b-naphthol) (1). < these No. 11 had been reported on contradictorily, and the remaining three had not been reported on at all. In addition to these the following three oil-soluble colors, not listed m the Green Tables and not reported on in literature, were wanted: 1. o-Toluidin-azo-b-naphthylamin (which does not seem to be described in litera- ture in any way). 2. Amidoazo-toluol. 3. Anilin-azo-b-naphthylainin. So that out of a total of seven oil-soluble colors on the United States market in the summer of 1907 only one had been examined physiologically and that with contradictory results. As before stated, these oil-soluble colors all belong to the class of nonsulpho- nated azo-colors. Fraenkel (p. 575) says: "When, however, the azo-colors contain no sulpho-group (i. o., are nonsulphonated) they are poisonous. Thus for example, Bismarck Brown * * * Sudan I." Yet meta-nitrazotin, a nonsulphonated color not in the Green Tables, and probably not upon the market anywhere, and which is meta- nitranilin-azo-b-naphthol, is according to Weyl nonpoisonons. Fraenkel (p. 575) also stated: The fact that the monazo colors examined by Cazeneuve and Lepine arc harmless, a? above stated, is equally explained by the constitution ot* these substances. Those two investigators examined (omitting the trade names) a-naphthylamin-sulphoacid- a/.o-a-naphthol-a-sulphoacid, a-naphthylainin-sulphoacid-a/.o- b- naphthol -disulpho- acid, a-naphthylamin-azo-b-naphthol-disulphoacid, xylitlin-a/.o-b-naphthol-disulpho- acid, Milphaiiilicarid-azo-a-naphtho], Amido-a-zo-toluene-disulphonicacid. These substances arc all sulphoacids and the sulpho-groupe here effect the d<'i>oi8on- ing of the original substance. Examining this statement it therefore appears that in Fraenkel's opinion, at least, a-naphthylamin-azo-a-naphtihol, a-naphthylamin- azo-b-naphthol, xylidin-azo-b-naphthol, anilin-azo-a-naphthol, and amido-azo-toluol are in and of themselves poisonous substances, and arc rendered nonpoisonons by snlphonation. Diligent search through the literature has failed to uncover any original communications of investigators to that effect. With hut one known exception all the nonsulphonated azo-colors egarded by Fraenkel as poisonous. The colon 17, IS, 41, 197, and 201 on page 159 are not used as oil-soluble colors and therefore form no comparative basis for judging oil-eoluble colors exeept that REASONS FOR SELECTING SEVEN PERMITTED COLORS. 161 both are nonsulphonated azo-colors; in FraenkeFs opinion, however, such a comparison is wholly justified and proper. Excluding now 17, 18, 41, 197, and 201 for the purposes of com- parison, the following tabulation is made: Oil-soluble nonsulphonated azo-colors on the market, 1907, deemed poisonous by Fraenhel. Oil-soluble colors. States calls market. poisonous. anilin-azo-resorcin x anilin-azo-b-naphthol x x xyii1°— Bull. J i: u —11 162 COAL-TAR COLORS USED IN FOOD PRODUCTS. Rule III. All coal-tar colors which have been examined physiologi- cally and have been declared to be of doubtful harmlessness shall not be permitted for use in foods. Rule IV. Only those coal-tar colors whose chemical composition was definitely disclosed or otherwise ascertained, and winch were on the United States market in the summer of 1907, and winch have been examined physiologically and with no other than a favorable result shall, for the present, be permitted for use in foods. ANALYSIS OF THREE RECOMMENDATIONS MADE TO THE DEPART- MENT OF AGRICULTURE. These rules were formulated as a guide in view of the divergent opinions expressed in three different recommendations to the Depart- ment of Agriculture. One of these recommendations suggested that permitted colors be designated by nine titles. Comparison of these titles with the Green Tables and with the tabulated survey of the unfa- vorable, favorable, and contradictory literature corresponding to the Green Table numbers (p. 63) discloses the following facts: Comparison of nine suggested color titles with corresponding Green Table numbers and the reports on the same. Name. Green Table No. Un- favor- able. Fa- vor- able. Con- tra- dic- tory. No re- port. 1 Name. Green Table No. Un- favor- able. Fa- vor- able. Con- tra- dic- tory. No re- port. 1 Chrysoidin 17 18 41 84 85 86 87 88 95 92 102 13 15 44 55 56 57 108 114 140 X X X X Ponceau 147 148 150 1()0 163 169 65 107 157 170 171 198 244 M 46a 1 X Tropueolin X X X X X X X X X X Blebrich Red (?).. 8 ii l p b o d a t e d X X X X X X X X X X X X X X X X Naphlhol Yellow S Total 36 8 13 15 1 Of physiological tests in literal lire Compiled. Therefore, under 9 titles 36 di Ferenl chemical individuals would be placed upon i lie permitted list , of which only 8, or less i ban 25 per cent, bave been examined physiologically with only favorable results, and 28, ot more I ban 75 per cent , had either not been examined at all physiologically or with contradictory results. REASONS FOR SELECTING SEVEN PEEMITTED COLORS. 163 The 8 of these chemical individuals examined xvith only favorable results, and the number of sources out of a possible 12 offering them on the United States market in the summer of 1907, are as follows: Colon reported on favorably and number of dealers handling same. Green Table No. Sources offering. Green Table No. Sources offering. M 65 86 92 10 2 2 102 107 169 462 i 2 1 The italicized Green Table numbers are those of the permitted list of Food Inspection Decision No. 76. Another recommendation suggested the permissive use of 42 entries in the Green Tables, which are tabulated below in the same manner as the suggestions in the preceding recommendation: Comparison of 42 recommended Green Table numbers with reports in the literature. Green table No. Un- favor- able. Favor- able. Equiv- ocal. No report. 1 Green table No. Un- favor- able. Favor- ' Equiv- able. ocal. No report. 1 4 8 17 50 53 54 55 65 85 86 89 92 94 LOT 103 107 138 tea 166 l's7 X 427 434 44s 451 456 457 462 477 478 479 480 504 512 513 514 516 517 5is 50'2 601 Total X X X ' x | X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X 5 14 15 8 1 Of phyriologtoal I In compiled literature. Applying the conclusions hereinbefore reached i<> these entries, it is found that 1 ! out of the 42 colors recommended, or exactly One-third, had been examined physiologically with only favorable results, and the remaining two-thirds had been either examined physiologically with only unfavorable of- with conflicting results, d not been examined at all. Hie M chemical individuals examined with favorable result and tin' number of BOUrcee OUl of a possible \2 offering them on the United State- market in the summer of 1907 are a- follows: 164 COAL-TAB COLORS USED IN FOOD PRODUCTS. Colors reported on favorably and number of dealers handling same. Green table No. Sources offering. Green table No. Sources oifering. U 65 85 89 92 102 103 10 2 2 1 6 107 166 240 462 477 512 577 1 2 3 E * The italicized Green Table numbers are those of the permitted list of Food Inspection Decision No. 76. The third of these recommendations suggested 27 different chemical individuals for permissive use; these are tabulated below in the same manner as the suggestions in the preceding recommendations. Twenty-seven colors recommended for use and reports in the literature on the same. Green table No. Un- favor- able. Favor- able. Equiv- ocal. No report. 1 Grern table No. Un- favor- able. Favor- able. Equiv- ocal. No reports 4 16 17 18 41 65 85 102 197 201 398 457 462 476 477 X 481 512 516 517 518 521 532 584 601 602 650 692 Total X X X X X X X X X X X X X X X X X X X X X X X X X 4 10 10 3 X 1 Of physiological tests in literature compiled. Therefore, applying the same method of drawing conclusions as in the case of the preceding two recommendations, it appears that 10 out of the 27 suggested chemical individuals had been examined physiologically with only favorable results, and the remaining 17 had either not been examined at all or with unfavorable or con- flicting results. The 10 chemical individuals examined with favorable results, and the number of sources out of a possible 12 offering them on the United States market in tho summer of 1907, are as follows: The 10 recommended colors favorably reported on in the literal are and the dealers handling same. Green Bouroei Green Souroea Table han- Table han- No. dling. No. dling. W id •177 2 6i a 8 M 2 r,n 6 KIJ :,'i 462 2 692 3 *Tho italicized Ghm-ii Table Qtimbexi in thoie tA the permitted list of Food Inspection Decision No. 76. REASONS FOR SELECTING SEVEN PERMITTED COLORS. 165 All of the colors included in these three recommendations con- cerning which only favorable reports were found in the literature are given in the following table : Recommended colors favorably reported on in the literature. Recommendations. Recommendations. Green Dealers Green Dealers Table otter- Table offer- No. I. ir. in. Total. ing. No. I. ii. III. Total. ing. U X X X 3 10 240 X 1 1 65 X X X 3 2 462 X X X 3 2 85 X X X 3 2 477 X X 2 89 X 1 1 512 X X 2 3 92 X X 2 617 X X 2 5 102 X X X 3 521 X 1 103 X 1 8 692 X 1 3 U07 166 169 2 1 1 7 1 X X Total . 8 14 10 i The italicized Green Table numbers are those of the permitted list of Food Inspection Decision No. 76. This table is now rearranged to show — I. The recommended colors not offered in the United States market in the summer of 1907. Green Table No. Recom- mended by- Green Table No. Recom- mended by- 92 102 166 2 3 1 477 521 2 1 II. The recommended colors offered on the United States market in the summer of 1907. Green Table No. Recom- mended by- Sources handling. Green Table No. Recom- mended by- Sources handling. M 65 86 89 103 107 3 3 3 1 1 2 10 2 2 1 168 240 612 617 692 l 1 3 2 2 1 1 2 3 5 3 i The italicized Green Table numbers an- those of the permitted list of Food Inspection Decision No. 76- It appears that out of 17 different chemical individuals suggested by these 3 recommendations jointly, 5 were not on the United Statei market in the summer of L 907, and L2 were on thai market; also that of these 12, 4 were suggested by all :; recommendations; 3 were wanted by only 2 out of llie .!. and the remaining 5 were spoken for by only one. Examining all of the recommendations, it. appears that. out. of a total of 79 recommendations 1 7 bad been examined with only favorable results, and that the remaining 62 were either not examined or had been examined with unfavorable or conflicting results. To add to the confusion the third recommendation mentioned sug- gested that the following ( oven Table numbers should not f>e permit- 166 COAL-TAR COLORS USED IX FOOD PRODUCTS. ted for use in food products because they are harmful : 1 ; 2 ; 3 ; 6 ; 86 ; 95; 425; 451; 454; 480; 483; 484; 487; 488; 490; 571; 584; 599; 620; 624; 650. The same paper, however, recommended that Nos, 584 and 650 be permitted for use in food products; that No. 86, sug- gested by both the other recommendations, be prohibited; and that Nos. 95, 451, and 480, suggested by one or the other of the remaining two recommendations, be likewise prohibited for use in food products. It must be clear to everyone that these three recommendations are not of one mind as to (1) the colors which are harmless; (2) the colors which are harmful; (3) vho colors which are used or are useful in food coloring; and (4) the colors which should be used in food coloring. Such a state of affairs fully justifies rules of the scope and intent of those formulated at the beginning of this section for making a selection of permissible food colors. PROCESS OF ELIMINATION. From the foregoing data it appears that out of 80 different chemical individuals on the food-color market of the United States in 1907, only 16 had been examined physiologically with a favorable result (see p. 64). These 16, arranged according to the shade produced, are as follows (numbers preceding the names are the Green Table numbers; the numbers in parentheses show the number of sources desiring the color; the italicized Green Table numbers are the ones finally selected by Food Inspection Decision No. 76) : 65. Fast Red B (Alphanaphthylamin azo-R-salt) (2). 103. Azorubin S (Naphthionic acid azo NW acid) (6). 105. Fast Red E (Naphthionic acid azo-Schaeffer acid) (1). 107. Amaranth (Naphthionic acid azo-R-salt) (7). 169. Crocein Scarlet 713 (Amidoazotolutol-monosulphonic-acid-azo-betanaphthol- monosulphonic acid B) (1). 240. Congo Red (Benzidin di.sazo-naphthionic acid) (1). 4G2. Acid Magenta (Magenta trisulphonic acid) (2). 512. Eosin A (Tetra-bromo-fluorescein) (3). 517. Erythrosin (Tetra-iodo-flnorcBcein) (5). 520. Rose Bengal (Dichlor-tetraiodo-fluorescein) (2). ORANGE. 85. Orange I (Sulphanilic arid azo-alpha-naphthol) (2). FELLOWS. 4. Naphthol Yellow (DinitroalphanaphtholBulpho acid) (10). 89. Brilliant Yellow S (Sulphanilic acid azo-diphcnylaniin, sulphonated) (1). 433. Guinea Green B (Benzyldehyde + benzylethyanilin Bulpho-acid) (1). .. Light Green 8 F yellowish (Beniyldehyde+benxylethylanilin-fiulphona- tton) (4). REASONS FOR SELECTING SEVEN PERMITTED COLORS. 167 BLUE. 692. Indigo Carmine (Indigo disulpho acid) (3). Considering now the reds, and particularly 107, which, as the num- ber appearing in parentheses after the scientific name indicates, was wanted by 7 sources out of the 12 drawn on, this being the most desired of all the reds, it would seem reasonable to believe that all the wants supplied by 65, 103, 105, 169, 240, and 462 would be covered by 107. Chemically 107 is closely allied to 65, 103, 105, and tinc- torially it is likewise closely allied to 169, 240, and 462. Nos. 512, 517, and 520 are chemically quite different from the other members of this group, and tinctorially they differ, being of a particu- larly brilliant shade, and tinctorially more powerful; 517 being desired by 5 out of the 12 sources drawn upon, and thus being the most desired of these three colors, was selected in the expectation that any work 512 and 520 could do in food products would be equally well done by 517. For reds, therefore, the choice fell upon 107 and 517. Orange. — Only one color was wanted, and that by 2 out of the 1 2 sources, and this was placed in the permitted list. Yellows. — The choice of No. 4, which was wanted by 10 out of the 12 sources, was made in the expectation that every purpose that 89 could serve in food products could be served also by 4. Greens. — Tinctorially and chemically the two greens are very closely allied, and in view of the reasonableness of the expectation that 435 could do all the work of 433 as well, its choice was regarded as justified, especially as 435 was wanted by 4 out of the 12 sources, as against only one for 433. Blue. — Xo. 692 being regarded as harmless by all, and being the only blue in the list, it was selected. In this manner six out of the seven permitted colore were selected. REASONS FOR ADDING PONCEAU 3R. An examination of the table on page 20 discloses the fact that among the reds desired four were azo-reds made from anilin deriva- tives as the first component, namely: 53. Xylidin-azo-alj>lmn:i|>hth<>lt. then, nor is it now, apparent, but the propriety of giving even a seeming need due consideration w as regarded as justified. For only one of the four desired, namelj, 55, could any speciiic references in the literature be 168 COAL-TAR COLORS USED IX FOOD PRODUCTS. found, and, in fact, four references were found favorable thereto and three unfavorable thereto; thus eliminating 55 from consideration, leaving 53, 54, and 56. The Austrian law of January 22, 1896 (see Lieber, p. 15), permits the use of azo colors derived from higher homologues of anilins, beginning with xylidin and sulphonated betanaphthols. As Nos. 53 and 54 were both derived from xylidin, it was regarded as safe to assume that they would probably be as objectionable as 55. This left 56 only to be considered, which color is derived from cumidin and R-salt. The cumidin portion satisfies the Austrian law and also satisfies the general law laid down in Fraenkel (p. 162), namely, that the greater the number of ring-methyls the less the toxic property of the resulting compound, because of the oxidation of these methyls to carboxyls in the animal system; the R-salt portion satisfies the gen- eral law laid down in Fraenkel, that the more highly sulphonated the less toxic a substance becomes, and therefore 56 is in the first portion of its composition of such a nature as to be less objectionable than 55, if 55 be objectionable. This combination of facts, namely, the desirability of an anilin azo-red, the provision for diminishing the toxicity of 55 by the sub- stitution of cumidin for xylidin therein, and the general provision in the Austrian food law, made it appear desirable and safe to take into account an anilin azo-red, and therefore to select 56 as probably the least objectionable, if it be at all objectionable, of the anilin azo-reds. The full list of permitted colors was therefore extended to 7, as follows : Red shades. — 107. Amaranth; 56. Ponceau 3R; 517. Erythrosin. Orange shade. — 85. Orange I. Yellow shade. — 4. Naphthol Yellow S. Green shade. — 435. Light Green S F yellowish. Blue shade. — 692. Indigo disulpho acid. It should be noted that with respect to tetra-iodo-fluorescein, 517 of the Green Tables, no specific investigation, pronouncing it harmless or harmful, is described in the literature. No. 516, the diodo-fluores- cein, is reported specifically adversely in the literature, especially by Chlopin; the Confectioners' List, the laws of Austria, the law r s of France, the rather superficial examination of No. 517 by Gran dho mine, and the statement in Fraenkel (p. 574) > that it produces no disturbance, are all taken to apply to 517 and not to 516. The identification of the color under examination with 517 has not in all cases been satisfac- torily exclusive, but it is believed that the differentiation of the harmful 516 by Chlopin makes all the other references cited pertinent to 517. LISTS OF COLORS RECOMMENDED BY INDIVIDUALS. 169 With respect to 435, it must be pointed out that 434, which is the methyl instead of the ethyl derivative, has only been regarded as suspicious, and one examination of 435, namely, that of Lieber (p. 144), does not appear to disclose anything which would positively exclude 435. QUALITY, CLEANLINESS, AND EFFICIENCY. The justification for limiting the permitted colors to 6 out of 16, against which nothing unfavorable is contained in the literature and regarding which favorable statements are at hand, and the addition to these 6 of the seventh color is to be found in the very great variation in tinctorial quality, in percentage of coloring matter, in amounts of insoluble matter, both organic and inorganic, and in the amounts of organic matter not coloring matter, as well as in the large number of samples containing an amount of arsenic in excess of that permitted by the United States Pharmacopoeia for the only coal-tar color therein mentioned, namely, Methylene Blue, and the varying amounts of heavy metals, such as copper, lead, and iron, mostly in excess of the limits permitted in the Pharmacopoeia for various medicinal chemicals. This wide variation in quality and degree of cleanliness, all pointing to a very great difference in the care with which coal-tar colors offered for food purposes are prepared, render the conclusion safe that some control over food colors in respect to quality is desir- able, necessary, and essential. The efficiency of the 7 colors, 6 of winch were selected from among the 16 considered for the purpose of making this list of permitted colors as being reasonably sure to be harmless, is evidenced by the fact that, although the addition of colors has been sought by persons interested in the food-coloring art, not one of the remaining 10 colors of those 16 has been so requested. In other words, the colors that the depart- ment has boon requested to add to the permitted list were outside of the 16 colors which were on the markets of the United States in 1907, and were described in the literature in such a manner as to lead to the conclusion that they were probably not harmful. XIII. LISTS OF COLORS SUBSEQUENTLY RECOMMENDED BY INDIVIDUALS AND ASSOCIATIONS. Since Food Inspection Decisions Nos. 7(> and 77 were published recommendations of lists of permitted colors have been made by other individuals and by a voluntary association. For the purpose of comparing these proposed lists of permissible colors with the per- mitted list of Pood [nspection Decision No. 76, the former are now to be examined in (he same manner as the coal-tar colors, on the United States market in the Bummer of 1907 for food coloring pur- 170 COAL-TAR COLORS USED IN FOOD PRODUCTS. poses, were examined in order to establish the permitted list of 7 colors. These recommendations cover the following lists: 1. W. Ernst's list. 2. Muttelet's interpretation of the French law. 3. Second International White Cross Congress list. 4. Beythien and Hempel's list. 5. An American manufacturer's list. 6. Belial's list. ERNST. W. Ernst (Fdrber Zeitung 1908, vol. 19, p. 381; abst. CTiem. Ztg. u Reportorium" 1909, p. 89) recommends for use in foods the following 38 titles of coal-tar colors. The Green Table numbers appear in parentheses after each title where corresponding numbers could be ascertained; the italicized numbers are those of the permitted list of Food Inspection Decision No. 76. 1. Auramin (425, 426). 21. Rhodamin (496, 497, 498, 502, 504, 2. Naphthol Yellow S (4). 505). 3. Quinolin Yellow (666, 667). 22. Acid Magenta (462). 4. Tartrazin (94). 23. Safranin (583, 584, 585). 5. Acid Yellow (8, 88, 95, 4). 24. The Croceins (13, 104, 106, 145, 151, 6. Spirit Yellow (Amidoazobenzol) (7). 160, 164, 169). 7. Curcumin S (399). 25. Acid Green (434, 435). 8. Eosin (512, 514, 515, 517, 521). 26. Brilliant Fulling Green (?). 9. Erythrosin (516, 517). 27. Malachite Green (427, 428). 10. Fluorescein (510). 28. Brilliant Green (428). 11. Orange II (86). 29. Water Blue (480). 12. Crocein Orange (13). 30. Patent Blue (440, 442). 13. Basic Oranges (?). 31. Brilliant Fulling Blue (?). 14. Fast Red (63, 65, 102, 103, 105, 107, 32. Domingo Blue B extra (?). 144). 33. Methyl Violet (451, 454). 15. Amara Red (?). 34. Acid Violet (464, 465, 467, 468, 470, 16. Naphthol Red (?). 471, 472, 474, 507). 17. Azo red (62). 35. Bismarck Brown (197, 201). 18. Bordeaux (65, 107, 157, 170, 171, 198, 36. Acid Brown (133, 138). 244). 37. Nigrosin (600,602). 19. Victoria Rubin (?). 38. Several Acid Blacks (184 and ?). 20. Ponceau (13, 15, 44, 55, 56, 57, 108, 114, 146, 147, 148, 150, 160, 163 L69, 448). Of these 38 titles only 11 refer to a single entry each in the Green Tables (namely 2, 4, 6, 7, 10, 11, 12, 17, 22, 28, 29); 7 titles can not be definitely connected with any entry in the Green Tables (titles 13 15, 16, 19, 26, 31, 32); one title is broader than the corresponding color in the Green Tables (title 38) and the remaining 19 titles each and all refer to more than one entry in the Green Tables. LISTS OF COLORS RECOMMENDED BY INDIVIDUALS. 171 These 31 titles embrace 88 different entries in the Green Tables; some of these entries are included in two or more titles as follows: Green Table number. Titles. 4 2,5 13 12,20,24 65 14, 18 107 14,18 Green Table number. Titles. 1G0 20,24 169 20,24 428 27,28 517 8,9 These 88 different Green Table entries can be divided as follows on the basis of the compilation of literature on physiological action here- inbefore given (see p. 63) : Unfavorable only.— 34, 164, 201, 425, 434, 502, 516, 602, 667. Total 9, or 10.2 per cent. Favorable only.— 4, 65, 102, 103, 105, 107, 169, 399, 435, 462, 467, 512, 517, 521, 600. Total 15, or 17.0 per cent. Conflicting— -8, 13, 15, 55, 86, 88, 95, 106, 138, 160, 163, 197, 427, 428, 448, 451, 480, 504, 584. Total 19, or 21.6 per cent. Not reported on.— -7 ', 44, 56 \ 57, 62, 63, 104, 108, 114, 133, 144, 145, 146, 147, 148, L50, 151, 157, 170, 171, 184, 198, 244, 426, 440, 442, 454, 464, 465, 468, 470, 471, 472, 174, 496, 497, 498, 505, 507, 510, 514, 515, 583, 585, 666. Total 45, or 51 per cent. According to this mode of judging only 15 ; or 17 per cent, of the colors suggested by Ernst for food coloring would be regarded as proper for use in foods. These 15 embrace 4 of the 7 permitted colors of Food Inspection -ion No. 76, namely, 4, 107, 435, and 517, leaving 11 to be con- sidered. Of these, 5 were not on the United States market in the nier of 1907, namely, 102, 399, 467, 521, and 600. The remaining 6 are as follows (the numbers in parentheses being the number of sources out of a possible 12 offering them on the United States market in the summer of 1907): (jo (2); 103 (6); 105 (1); 169 (I); 462 (2); and 512 (3). The reasons for the noninclusion of these in t ho permitted list of 1 Inspection Decision Xo. 76 have been given on page 167. MTJTTELET'S INTERPRETATION OF THE FRENCH LAW. Muttelet (Annales des Falsifications, 1909, pp. . places the ; Mowing interpretation on the French regulations <>f December 29, 0, and of August t. L908, classifying them as — I. Those colon irhich are certainly permitted by those regulations. il. Those <«)l<.rs whose permitted or prohibited use i.-> doubtful. III. Those colon which arc certainly prohibited. The Green Table numbers contained in each class are classified as follows (the italicized numbers being those of the permitted list of i Inspection Decision No. 76): (a) Unfavorably reported; (5) iSe*- i mis r.r Inotadlnc Pi I List. 172 COAL-TAR COLORS USED IN FOOD PRODUCTS. favorably reported; (c) conflictingly reported; (d) not reported in the literature as to their physiological action (see p. 63) ; in paren-. theses is given the number of sources out of a possible 12 offering those colors on the United States market in the summer of 1907. I. Those colors which are certainly permitted: (a) None. (b) 4 (10); 5 (0); 462 (2); 512 (3); 517 (5); 520 (2); 600 (0); total, 7. (c) 55 (2); 427 (2); 451 (5); 457 (0); total, 4. (d) 158 (0); 518 (2); total, 2. II. Those colors whose permitted or prohibited use is doubtful: (a) None. (6) 65 (2); 107 (7); 599 (0); total, 3. (c) 8 (5); 84 (2); 427 (2); 457 (0); 601 (1); total, 4. (d) 452 (2); 456 (0); 513 (0); 514 (0); total, 4. III. Those colors which are certainly prohibited: (a) 3 (0); (b) none; (c) 427 (2); (d) none. It will be noted that 427 appears in all three of Muttelet's classifi- cations. The reason for this is that 427 is or has been marketed in at least three different forms; of these the straight chlorhydrate is per- mitted, the oxalate and the zinc chlorid double salt are of doubtful admissibility, and the picrate is undoubtedly forbidden. Also 457 appears in Muttelet's Classes I and II because the "Bleu Lumiere" of Muttelet's Class I is indistinguishable from his "Bleu Lumiere" of Class II, when using the Green Tables as a guide. Fur- ther, this list of Muttelet contains only 3 out of the 7 colors permitted in Food Inspection Decision No. 76, namely, 4 and 517 of his Class I and 107 of his Class II. An inspection of Muttelet's Class I discloses 13 Green Table entries, of which only 7 have been reported in the literature, as herein compiled and rated, in a favorable manner; of these 7, 2 were not on the United States market in the summer of 1907 and of the remaining 5, 2 are in the permitted list of Food Inspection Decision No. 76; the remaining 3 are: 462. Acid Magenta (2); 512. Eosin (3); 520. Rose Bengal (2), and the reasons for whose noninclusion in the permitted list of Food Inspection Decision No. 76 have been given (p. 167). SECOND INTERNATIONAL WHITE CROSS CONGRESS. The Second International White Cross Congress, held in Paris, October 18 to 24, 1900, according to the Chemiker Zeitung, 1909, page 1227, adopted the following list of colors which were said to be proper for use in coloring food products. Tho figures in parentheses are tho Green Table numbers; the italicized numbers are those of the permitted list of Food Inspection Decision \o. 76. 5. Fasl Red E (106). 6. New Coccin (10(5). 7. Ponceau 2R (66). 1. Erythronn (616, 517). J. lUiodaniin B (504). 3. Bordeaux S (107). 4. Bordeaux G (170). 8. Xylidin Scarlet (55). LISTS OF COLORS RECOMMENDED BY INDIVIDUALS. 173 9. Magenta (448). 10. Acid Magenta (462). 11. Orange I (85). 12. Naphthol Yellow S (4). 13. Chrysoin (84). 14. Auramin (425). 15. Acid Green (434, 435). 16. Lyons Blue (457). 17. Patent Blue (440, 442). 18. Paris Violet (451). 19. Acid Violet (464, 465, 467, 468, 470, 471, 472, 474, 507). 20. Black Indulins (599). 21. Sulphonated Nigrosin (602). It will be observed that titles 7 and 8 refer to the same Green Table number; this leaves, therefore, only 20 titles to consider. These refer to 31 different Green Table numbers which are classified as (a) only unfavorable reports, (b) only favorable reports, (c) conflicting reports, and (d) no reports, in the literature hereinbefore compiled and rated (see p. 63) : (a) 425, 434, 516, 602; total, 4. (b) 4, 85, 105, 107, 435, 462, 467, 517, 599; total, 9. (c) 55, 84, 106, 448, 451, 457, 504; total, 7. (d) 170. 440, 442, 464, 465, 468, 470, 471, 472, 474, 507; total, 11. The numbers of section (b), of which only favorable reports are recorded, are the only ones here of interest; they are 9 in number, or less than 30 per cent of all those included in tins list, and of these 9, 5 are on the permitted list of Food Inspection Decision No. 76. The remaining 4 are the following, the number in parentheses represent- ing the number of sources out of a possible 12 offering them on the United States market in the summer of 1907: 105. Fast Red E (1). 462. Acid Magenta (2). 467. Acid. Violet 6 B, not offered. 599. Printing Blue, not offered. The reasons for the noninclusion of these colors in the permitted list of Food Inspection Decision No. 76 have been given on page 167. BEYTHXEN AND HEMPEL. Beythien and Hempel (Farber Ztg., 1909, v. 15, pp. 301, 348, 392, 436; abst. Ohem. Ztg., 1910, p. 58) recommend the following colors for use in food products. (The numbers in parentheses arc the corre- Bponding Green Table numbers, where Buch connection could be established; the italicized numbers are those of the permitted list of Food Inspection Decision No. 76.) l. AJiaarinblue (682, 5G:J). l». Amaranth {107). :;. Bordeaux Red (?). l. Brilliant Blw 5. Diamondfacfa i Blue ( • 898 MO). fellow It (9). I Red (68, 86, 102, 103, LO 512, 51 i. 515, 517, 521). 116, 517). n. FuchaiD - 12. Light Green 8 V yellowiah (435). IS. Indigo diaulphoacid • li. [nduli : :ht Blue it;. Malachite Green 17. Methyl Violet (451, 4 174 COAL-TAR COLORS USED IN FOOD PRODUCTS. 18. Xaphthol Yellow S (4). 26. Roccellin (102). 19. Orange I (85). 27. Roscellin (?). 20. Orange L (54). 28. Rubin (448). 21. Paris Violet (451). 29. Acid Yellow S (4). 22. Phloxin (518, 521). 30. Acid Magenta (462). 23. Ponceau 3R (56, 57). 31. Solid Blue (?). 24. PoDceau Red (?). 32. Tropaeolin OOO (85, 86). 25. Primrose (3, 513, 514). 33. Waterblue (480). For six titles (3, 4, 15, 24, 27, and 31) no corresponding Green Table number could be determined. Of the remaining 27 titles, Xo. 9 includes part of No. 10, 32 includes 19, 9 includes part of 25, 8 includes 26, 17 includes 21, 11 is identical with 30, and 18 is identical with 29, all on the assumption that the correct connections between title and Green Table numbers have been made. Therefore these 27 titles are, in fact, only 25 titles; of these 25 titles, 13 refer to but one Green Table number each; the remaining 12 titles each refer to two or more Green Table numbers. The Green Table numbers above given are now arranged in the following four classes according to the literature hereinbefore com- piled and rated (see p. 63) : (a) Only unfavorable reports: 3, 516, 639; total, 3. (b) Only favorable reports: 4, 65, 85, 102, 103, 105, 107, 435, 462, 477, 512, 517, 521, 599, 692; total, 15. (c) Conflicting reports: 9, 86, 427, 428, 448, 451, 480, 563, 601; total, 9. (d) No reports: 54, 56, 1 57, 63, 144, 454, 513, 514, 515, 518, 562, 603, 640; total, 13. The 15 Green Table numbers of class (b) are the only ones here of interest; they include 6 out of the 7 permitted colors of Food Inspec- tion Decision 76; of the remaining 9, 4 (102, 477, 521, 599) were not on the United States market in the summer of 1907. The remaining 5 are as follows. (The number in parentheses is the number of sources out of a possible 12 offering them on the United States market in the summer of 1907): 65. Fast Red B (2). 462. Acid Magenta (2). 103. Azorubin S (6). 512. Eosin (3). L05. Fast Red E (1). The reasons for the noninclusion of these colors in the permitted list of Food Inspection Decision Xo. 76 have been given on page 167. SUMMARY OF THREE PRECEDING RECOMMENDATIONS. The recommendations made by Ernst, the White Gross Congress, and Beythien and Hempel are summarized in the following table: i if permitting qm of Ponceau SB, LISTS OF COLOKS RECOMMENDED BY INDIVIDUALS. 175 Summary of recommendations from three sources. Green Table No. (11). Unfavorable reports only. Number recom- mending. Dealers offering. Ernst. White Cross. Beythien and Hempel. 3 x 1 6 2 3 1 -' 1 i) 1 94 104 201 425 434 502 510 002 039 X X X X X X X X 1 1 1 2 2 1 3 2 1 X X X X X X 007 Total . . X 9 4 3 Green Table No. (19). Favorable reports only. Number recom- mending. Dealers offering. Ernst. White Cross. Beythien and Hempel. 4 05 85 X X X X X X X X X X 3 2 2 2 2 3 3 3 3 2 1 2 3 2 2 1 1 10 2 2 1 7 1 4 2 3 5 3 X 102 103 105 107 109 399 435 482 467 477 X X X X X X X X X X X X X X X X X X X X X 512 517 521 599 X X X X X 000 692 X X Total.. 15 9 15 Green Tab!.- No. (21). Conflicting reports. Numl>er recom- mending. I><':i!it-; oilering. Ernst. White Cross. Beythien ami Hempel. 8 x l l 1 1 l ■_> l l l l l l •j 3 I 1 5 1 2 H I 4 2 3 4 5 1 5 1 X 13 16 5.-. VI X X X X 1 86 106 138 100 183 197 4J7 • 461. X X X X X X X X X \ X X X \ X X X 668. X X 684. 0<)1 X I'l 7 9 176 COAL-TAR COLORS USED IN FOOD PRODUCTS. Summary of recommendations from three sources — Continued. Green Table No. (51). Not reported on. Number recom- mending. Dealers offering. Ernst. White Cross. Beythien and Ilempel. 7 44 54 X X 2 2 2 2 2 2 2 2 2 1 1 1 1 2 1 2 2 1 1 1 1 1 1 1 1 2 2 1 1 4 2 X X X 56 57 62 63 104 108 114 133 144 145 146 147 148 150 151 157 170 171 184 198 244 426 440 442 454 464 465 468 470 471 472 474 496 497 498 505 *07 510 513 X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X 514 515 518.. . X X 562 583 586 603 X X X X 840.. . 866 Total.. X 1 45 11 13 r These three recommendations embrace 105 different Green Table numbers. The following table discloses the conformity of the com- posite of these three recommendations to the United States market in lite summer of 1907: Number of suggested samples in the three supplementary lists found on the market, 1907. Data 1 in on 1 | States market. l'n favorable 11 l'.t M .M 7 13 17 in Total 105 47 LISTS OF COLORS RECOMMENDED BY INDIVIDUALS. 177 UNPUBLISHED RECOMMENDATIONS OF A MANUFACTURER. In addition to these published recommended lists the following recommendation has been made by letter by a manufacturer who contributed specimens of food colors to the United States market in the summer of 1907: That for the following permitted colors of Food Inspection Decision No. 76, there be substituted certain colors, the Green Table numbers alone being here given : rmitted. Substitute 4 94 56 53 85 15 517 521 In regard to these proposed substitutes it is to be said that Xos. 15 and 521 were not on the United States market in the summer of 1907, according to the canvass made and described in Section I, Nos. 53 and 94 were on that market, No. 53 had not been examined physiologically, and No. 94 had been examined physiologically with only unfavorable results. The way was, therefore, not open to placing any of these colors on the permitted list under the procedure adopted. However, had No. 15 been on the market it would not have been placed on the permitted list, because it has been examined physiologically with contradictory results, while No. 521 might have been placed on the permitted list because it seems to have been examined physiologically and with only favorable results. BEHAL. As the result of careful investigation Belial {Revue Generale des Matieres Colorantes, 1010, p. 131) suggests the use of 21 definite chemical individuals which are given in the following. With the tabulation on page 63 as a guide these 21 colors are classified as unfavorable, favorable, contradictory, and not reported on. The italicized figures are in the permitted list of Food Inspection Decision No. 76. UNFAVORABLE (1). 425. Auramin (S). FAVORABLE (11). Naphthol Xell<) specimens would disclose the qualitative conditions of the market so far as these seven permitted colors were concerned, and it was expected thai certain limits for standards of cleanliness would he fixed by such an examination. 182 COAL-TAR COLORS USED IN FOOD PRODUCTS. In examining these 30 specimens of the seven permitted colors chemically the following determinations were made: 1. Moisture. 2. Chlorin as chloride. 3. Sulphated ash, together with its iron, aluminum, calcium, and copper content, and the determination of the sulphuric acid in the sulphated ash. 4. Total sulphur. 5. Gutzeit test (test 17 of the United States Pharmacopoeia, eighth revision). 6. Heavy metals test (test 121 of the United States Pharmacopoeia). 7. Total insolubles, together -with the determination of the proportion that is vola- tile on ignition. 8. Ether extractive. The methods of analysis actually used on these 30 dye specimens are here given solely for the purpose of comparing them with the methods developed therefrom and presented beginning with page 210. Experience has shown that the methods here given are defective in many particulars, and therefore they are not to be used for exact work. MOISTURE. Dry a sample of each color weighing 3 grams at 105° to 108° C. for two hours. The loss in weight is assumed to be moisture. This method is not wholly accurate in the case of Naphthol Yellow, nor is it accu- rate in the case of Amaranth; but the scarcity of material made it seem unwise, at this stage, to undertake any extended investigation as to the amount or nature of the heating required surely to expel all moisture. The results, therefore, while not as accurate as might be desired, are, for the purposes of this exploratory investigation, sufficiently accurate for the object for which they were undertaken. CHLORIN AS CHLORIDS. Gently heat samples weighing 0.1 gram with 2 grams of sodium carbonate, and after destroying the greater part of the organic matter add 0.1 gram of powdered potassium nitrate and gently heat the whole until the organic matter is entirely destroyed. After cooling treat the whole with small amounts of cold water and remove from the crucible; effect the solution of the whole by gently heating. After cooling bring the bulk to about 150 cc, cool to room temperature, slightly acidify with nitric acid, pre- cipitate the chlorin with silver nitrate, and weigh as silver chlorid. SULPHATED ASH. Moisten half-gram samples with concentrated sulphuric acid, gently evaporate to dryness, and treat, the residue with 5 cc of concentrated sulphuric acid and again evaporate to dryness; repeat the operation until a white ash results, when the whole is ignited 1<> constant weight. After weighing take up the sulphated ash in boiling water (and if necessary, any undissolved material can be taken up with hydrochloric acid by treatment on a boiling-water bath.) Mix the two solutions and bring to a total volume of 200 cc. In one half thereof determine the contained sulphur by a precipitation with barium chlorid, and weigh BS barium sulphate; in the other half determine iron and aluminum by precipitation with ammonia, ignite and weigh ;is ferric and aluminum oxid. No separation of any contained aluminum was under- ANALYSIS OF PERMITTED COLORS, 1907. 183 taken in the nitrate from the iron; the lime was precipitated as oxalate. In the only rase that copper was encountered it was precipitated as sulphid before taking out the iron or the calcium. TOTAL SULPHUR. Mix 0.2 gram of the sample with 4 grams of sodium carbonate and 0.5 gram of potassium nitrate, and ignite to complete the destruction of the organic matter; take up in hot water; acidify with hydrochloric acid and precipitate the sulphur as barium sulphate with barium chlorid. GUTZEIT TEST. Mix 2 grams of the substance with 2 grams of a mixture of 1 part of potassium nitrate and 5 parts of sodium carbonate and ash in a porcelain crucible over a low flame; if not white when cool, mix the ash with 1 gram of potassium nitrate and again ash over a low flame. Generally the second ashing is sufficient. Dissolve the residue in 50 cc of hot water, boil, filter, neutralize with dilute sulphuric acid, and evaporate to substantial dryness on a boiling-water bath. Then add 1 cc of concentrated sulphuric acid and dry over a Bunsen flame; take up the residue with 5 cc of distilled water containing 0.5 cc of concentrated sulphuric acid and 10 cc of a saturated solution of sulphurous acid; evaporate the whole to a bulk of 5 cc on a water bath; add 20 cc of 8 per cent hydrochloric acid and subject this material to the action of 2 grams of metallic zinc, free from arsenic, which has been so activized by means of platinic chlorid that at the end of two hours more than 1 gram of zinc has dissolved and the evolution of gas has been constant and continuous. Conduct the reaction in a flask of 60 cc capacity with a neck 1 cm in diameter and 6 cm long. After the introduction of the solution and the zinc into the flask, stopper the neck of the bottle with gauze, the lower half of which is dry and the upper half moistened with the test solution of lead acetate of the United States Pharmacopoeia. After carefully wiping the lip of the fla'sk, cover it with Schleicher & Schiill quantitative filter paper which has been saturated three times with alcoholic mercuric chlorid solution, with complete drying between each saturation on one and the same spot of the filter paper. In the case of Naphthol Yellow S it is necessary to heat gently with 10 grams of sodium carbonate until the organic matter is substantially all destroyed, then add 1.5 grams of potassium nitrate and heat to complete destruction of the organic matter. Dissolve the fused mass in hot water and a few drops of fifth-normal sulphuric acid; then make distinctly acid with the same and add an excess of 1 cc of concentrated sulphuric acid; evaporate first on the water bath and afterwards on asbestos until all odor of nitrous fumes and of hydrochloric acid has disappeared. Take up the residue in 5 CC of water and 15 cc of a saturated solution of sulphurous acid in water. Beat the whole on the water bath until no odor of sulphur dioxid remains. The preparation of the sample by ignition with carbonate and nitrate should not bedone in platinum, butshould be done in a porcelain crucible, since it has hap; thai as much as 0.05 mg of arsenic mixed with the dye, which had been fused in platinum with carbonate and nitrate, could not be detected on the mercury-chlorid paper, whereas 0.01 mg when similarly treated in porcelain could always DO detected, and 0.006 mg would usually be found when done in porcelain. HEAY1 ICBTALS. Mix as much Oi the BUDfltance as approximately contains 1 gram «'f color with 10 times its we i -ht of carbonate of soda and Ignite with the addition tnoi potas- sium nitrate. Dissolve the whole in water, any u: 1 being taken up 184 COAL-TAR COLORS USED IX FOOD PRODUCTS. with hydrochloric acid, bring the two solutions together and slightly acidify with hydrochloric acid; dilute to 100 cc; place 10 cc in a test tube of 40 cc capacity and warm to 50° C. in a water bath; add 10 cc of a freshly prepared saturated solution of hydrogen sulphid in water, stopper the test tube well and allow the whole to stand in water having a temperature of 35° C. for a half hour. Dilute a 3.3 cc portion to 10 cc, treat as before with 10 cc of hydrogen sulphid in water for one-half hour at 35° C, and add to each ammonium hydrate. TOTAL INSOLUBLES. Dissolve 1 gram of the substance in 1 liter of water, filter through counterpoised quantitative filters, and wash with hot water until all traces of color are removed from the filter paper, then dry at 100° C. to constant weight and weigh; report the weight as total insolubles. Ignite the total insolubles in a platinum crucible and report the residue remaining as nonvolatile insolubles. ETHER EXTRACTTVE. Two methods were employed, the one consisting in direct extraction of the sub- stance in a Soxhlet apparatus by means of redistilled ether dried over sodium. Schleicher & Schull extraction cartridges were used after they had been thoroughly extracted by ether and shown by examination that they yielded no extractive to ether whatever. This is a necessary precaution, because the amount of ether extrac- tive matter in these cartridges varies. The amount of ether extractive was determined by driving off the ether over a 32-candlepower incandescent lamp, finally drying and cooling in a desiccator. This method, however, is not satisfactory, since the results it gives are undoubtedly low and it seems that the higher the material was dried the more erratic were the results. Therefore the following method was used: Disolve 1 gram of the sample in 100 cc of water; add 0.5 gram of sodium acetate; extract three times with 50 cc of ether in aseparatory funnel; mix the ether so recov- ered with 10 cc of water; separate the ether and dry with fused calcium chlorid; let stand from 12 to 24 hours; pour off from the calcium chlorid; distil off the ether as in the preceding method; acidify the color solution containing sodium acetate with 1 cc concentrated hydrochloric acid; extract three times with 50 cc of ether, and proceed as before. RESULTS OF CHEMICAL EXAMINATION, 1907. DETAILED CHEMICAL DATA OX EACH PERMITTED COLOR. The results of this examination are given in the following tabula- tions under the respective Green Table numbers; the serial num- bers refer to the numbers assigned to the specimens as they were received. The abbreviations "p." and "n. p." opposite "Gutzeit test," and "Heavy metals test" stand for "pass," and "not pass," respectively, according as the specimen did or did not comply with those requirements; the entries opposite "Ratio," indicate the degree of concordance of the sulphated ash as weighed, and tho contained SO, calculated back to sodium sulphate and is a measure of tho accu- racy or dependability of the sulphated ash item audits determination. ANALYSIS OF PERMITTED COLORS, 1907. 185 As a guide in ascertaining the conformity of these colors to their sup- posed standards when judged by the analytical data obtained, the following table may of service: Percentages of sulphur, sodium, and sulphated ash properly belonging to each of the seven permitted colors. Green Table No. 4 56 85 107 435 517 692 Name of colors. Sulphur. Naphthol Yellow Ponceau 3 R Orange I Amaranth... Light Green Erythrosin Indigotin 8.95 12. 97 9.16 15. 92 11.12 (') 13.75 Sodium. 1 Iodin in No. 5Vt .7 per c?nt. 12.87 9. 33 6.58 11.44 8.31 5.24 Sul- phated ash. 39.73 20.31 35.32 25.64 16.17 30.52 Detailed analytical data obtained on several samples of each of the permitted colors (per cent). GREEN TABLE NO. 4. NAPIITIIOL YELLOW S. Determinations. Moisture NaCl Na as NaCl Sulphated ash: Al and Ee Ca 'h S0 3 inNaiS0 4 =Na 2 S0 1 Ratio Na as Na2So< Total sulphur...'. Gutzeit test Heavy metals test Total insolubles Nonvolatile Volatile te ether extract.. Add ■ t Total ether extract Ether extract solid Serial Nos. 23. 0.70 6.73 2.65 .07 04 38. 00 38.18 100.3 12. 32 8.72 n.p. n.p. .05 ..03 .34 1. 32 .062 o. 28 3.47 1.37 .07 None. 48.8 49.57 101.6 15.82 9.47 n.p. n.p. None. None. .15 .90 LOS .042 1.' .' .40 .07 None. 47 1 47. 32 100.5 15.46 12. (.3 n.p. n.p. None. None. .14 .70 .84 .04 0.60 .61 .24 .07 None. 5L 4 51.76 100.7 16.68 12. 42 n.p. n.p. None. None. None. .16 .75 .91 .03 187. 0.33 2.86 1.13 .28 None. 4:'. 44 100.5 15. 95 n.p. n.p. .05 None. .05 .10 .32 Determinations. 201. 209. 228. 272. 280. 0.5C 9.31 .21 .03 49.20 Ml 1 n.p. n.p. . ID . 1 1 .47 .10 otao 1.61 101.3 i: 22 n.p. n. p. Ml .OCH 1.13 .07 Ml s n.p. n.p. . h .10 .80 .13 2. 12 L28 .07 N011.'. 45. BO 100.4 11 78 10.50 I'- ii. 1-. .'.HI None. u . 1 ■ .82 .01 Bulpl \i and Fe .07 40 11 . ulphur £6 Hea\ \ • Volatile in Acid etl Tot ll r'll.T i-.\tr:icl EihtT extract nlid 186 COAL-TAR COLORS USED IX FOOD PRODUCTS. Detailed analytical data obtained on several samples of each of the permitted colors (per cent) — Cont inued . GREEN TABLE NO. 56. PONCEAU 3R. Determinations. Serial No. 9. Determinations. Serial No. 9. 5.76 19.89 7.84 Ratio 101.20 NaCl .. Na as Na<>S0 4 12.36 Na as NaCl P- n.p. .28 Sulphated ash: Al and Fe | Heavy metals test .28 None. 38.10 38.56 Ca .83 Na 2 S0 4 Total ether extract 1. 11 S0 3 in Na 2 S0 4 -Na2S0 4 Ether extract solid .326 GREEN TABLE No. 85. ORANGE I. Determinations. Serial Nos. Determinations. Serial Nos. 16. 224. 16. 224. 11.06 8.78 3.18 .14 None. 30.80 31.18 101.3 9.99 5.39 3.51 1.38 .07 .14 18.00 18.09 100. 50 5.84 Total sulphur 10.12 P- P- 1.50 .20 1.30 .02 .20 .22 .214 8.34 NaCl ... n.p. P- .55 Na as NaCl. .. Sulphate ash: Al and Fe .20 Ca Volatile .35 Na? SO4 . .62 S0 3 in Na2S04-NajS0 4 .. .20 Ratio Total ether extract .82 Na as Na2S0 4 Ether extract solid . 254 GREEN TABLE No. 435. LIGHT GREEN SF YELLOWISH. Determinations. Serial Nos. 57. 92. 168. 233. 8.32 .041 .016 .70 3.29 9.00 7.78 86.46 2.D2 11.48 Tr.(p.) P- .20 None. .20 .02 .05 .07 .026 5.08 .102 .040 .28 .00 43.40 43. 96 101.30 14.07 1.-. 11 Tr.(p.) P- .45 .06 .39 .02 .05 .07 None. 5.04 1.735 .683 .14 .085 48.00 46.68 95.92 15. 56 15. 82 llea\ V. P- .75 .60 .15 .05 .05 .10 .02 4.15 NaCl 1.530 Na as NaCl .603 Sulphate ash: Al and Fe .14 Ca . .085 Na 2 S0 4 S3. 00 SOt in Na2S0 4 — Na 2S0 4 S3. 82 Ratio 101.3 Na as Na2SO« Total sul phur 17.19 16.75 Heavy. p. Total Insoluble! Nonvolatile .- Volatile .95 .60 .35 None. . 06 .05 .02 ORE E N TABLE No 107. AM \l: Will. Determinations. Serial NOS. 06. 130. 162. 177. no. 1 M 1 12 . It .215 65.20 6 16 .07 . 867 - 35. 79 ll in . it . 03 60 30 g si III (X) Us is 90 10. 52 4. 20 11. 21 . 12 00 00 61 26 8.24 24.21 v. M Sulphati m and Fe.. . .21 .57 vi BO 60. 72 ANALYSIS OF PERMITTED COLORS, 1907. 187 Detailed analytical data obtained on several samples of each of the permitted colors (per cent) — Continued. GREEN TABLE No. 107. AMARANTH-Continued. Determinations. Serial Nos. 82. Ratio 102. 00 Na as Na 2 SO< 21. 13 Total sulphur : 8. 62 Copper None. Gutzeit test. Heavy metals test Total insolubles Nonvolatile Volatile Acetate ether extract. Acid ether extract Total ether contract.. Ether extract solid... n.p. n.p. .30 .12 .18 .08 .07 .15 96. 130. 162. 100.80 17.22 13.02 None. n.p. n.p. 1.50 1.10 .40 .15 .15 .30 .062 100.80 19.39 8.57 None. n.p. n.p. 2.05 1.10 .95 .18 .15 .33 .052 101.5 15.82 9.42 .09 n.p. n.p. 1.05 .35 .70 .20 .03 .23 .040 177. 102.1 19.45 11.09 None. n.p. n.p. .90 .35 .55 .02 .03 .05 .044 219. 101.9 16.14 8.54 None. n.p. n.p. 2.40 1.15 1.25 .30 .10 .40 .054 GREEN TABLE NO. 517. ERYTIIROSTN. Determinations. Serial Nos. 184. 200. 216. 254. 9.6 NaCI (not determined) Sulphated a-'h: 0.14 None. 39.00 39.44 12.65 P- P- .00 .04 .50 .80 0.14 None. 35. 69 35.44 11.52 n.p. P- .20 None. .20 .35 .14 None. 31.80 3246 10.40 P- P. . 66 .20 .45 .10 ">* Ca NajSO< S0 3 in NajSO«=NajS0 4 Na as Na 2 SO< 26. 8 8.56 Gutzeit t^st P- P- .90 Total insolubles Nonvolatile .20 Volatile . 70 Acetate other extract . 10 Total ethrr extrnrt Ether extract solid .024 .07 .02 .02 GREEN TABLE NO. 692. INDIGO DISULPIIO ACID. Doterminations. Serial Nos. Determinations. Serial No*. 90. 195. 249. 90. l".-,. 249. Moisture 7.25 7.08 2.77 .42 2m. 74 5.32 11.26 .14 55.40 17. '7 7.31 7.02 2. 77 .71 102 B Total sulphur 10.30 P (?) LOB .09 .07 .008 .80 .if.' .10 ,U .OH 10m V.i l. 1. 15 \;i('l Sulphated ash: Heavy metals tost Total Insoluble* \i a 1 Pe - Acetate eth< • \ciii ether extract 1 ol il ether • Ether < 1 n> In NaiSO«=Na, I .01 RECALCULATION OF ANALYTICAL DATA ON BASIS <»l" COLORING KATTEB PRESENT. The difficulties in the way of translating these analytical data into proximate constituents are so great, in bo many of the cases, i make any attempts to obtain practical results in thai way absolutely 188 COAL-TAR COLORS USED IN FOOD PRODUCTS. useless; the amount of material on hand was unfortunately so small at the beginning (2 ounces or less, in most instances) that the utmost economy of material was necessary to get the data reported, on account of the large amount of material needed in the exploratory work done in trying out the methods for determining arsenic, heavy metals, and ether extractives. In the arsenic test the coloring matter, as a whole, was considered, and no attempt was made to get at the actual amount of real coloring matter in the exact weights of the material as a whole, used for such examination, and consequently the results are not translatable into actual weights of real color used. It is obvious that if 2 grams, or other weight of the substance as a whole, failed to pass any test when the whole amount was considered as though it were all color, the material could not possibly have passed those tests if amounts thereof, corresponding to the prescribed weights of actual color used, were taken; also it is clear that if the sample, as a whole, passed a given test, at a given weight thereof, it would not necessarily have passed that test had an amount thereof, corresponding to that same weight of actual color, been used. The amount of actual or real coloring matter in 27 of these 30 specimens is not greater than shown in the following: Per cent of actual coloring matter in 27 samples. Name of color. Serial No. Percent. Serial No. Per cent. Naphthol Yellow S Ponceau 3R Orange I Amaranth Light Green SF Yellowish Erythrosin Indigo Disulpho Acid 23 73 108 142 187 9 16 82 90 130 57 92 210 90 195 91.20 95.20 97.80 97.88 90.34 73.24 78.44 50.40 78.95 55.00 91.37 94.30 89.05 84.11 05.10 201 209 228 272 280 75.56 93.19 95. 72 92.81 79. 47 224 89. n 162 63.52 177 66. 40 219 64.26 168 92. 37 233 93.32 249 84.47 In order to make these analytical results comparable among them- selves and with other analytical results later to be given, the items common salt, volatile and nonvolatile insolubles, total insolubles, acetate ether extract, acid ether extract, and total ether extract of the preceding tabulations have been recalculated in parts per 100 of actual or real coloring matter as just enumerated: ANALYSIS OF PERMITTED COLORS, 190^ 189 Analytical data recalculated to parts per hundred of coloring matter present. GREEN TABLE NO. 4. NAPHTHOL YELLOW S. Determinations. Serial Nos. 23. 73. 108. 142. IS, 201. 209. 228. 272. 280. Common salt 7.38 3.64 1.04 0.62 2.97 31.53 4.38 2.99 3.51 11.82 Insolubles: Volatile .02 .03 .00 .00 .00 .00 .00 .00 .05 .00 .13 .27 .86 .00 .02 .10 .97 .00 1.46 .05 Total .05 .00 .00 .00 .05 .40 .86 .12 .97 1.51 Ether e-v- tract: .37 1.07 .16 .95 .14 .72 .16 .77 .10 .33 .19 .44 .27 1.16 .08 .94 .16 .78 .13 .97 Total 1.44 1.11 .86 .93 .43 .63 1.43 1.02 .94 1.10 GREEN TABLE NO. 56. PONCEAU 3R. Determinations. Serial No. 9. Determinations. Serial No. 9. 27.16 Ether ex tract : Acetate 0.38 Volatile Acid 1.13 Total 1.51 GREEN TABLE NO. 85. ORANGE I. Determinations. Common salt. Insoluble: \ "lutile Nonvolatile. Total Serial Nos. 16. 224. 11.19 3.91 1.66 . 25 .39 .23 1.91 .62 Determinations. Ether extract: Acetate. . . Acid Total. Serial Nos. In. 0.02 .25 .27 224. 0.69 .23 .92 GREEN TABLE NO. 107. AMARANTH. Determinations. Serial Nos. 82, 96. 130. 162. 177. 219. 84.87 39.97 42.85 37 39 1 Volatile .21 .51 1 1 1.71 1.10 1 93 1 78 Tot:il .53 1.90 3.69 1.65 1.37 3 71 Ether extract: .11 .12 .19 .19 ,1? .31 .03 46 And 15 Total .26 .38 .59 .36 .08 61 190 COAL-TAR COLORS USED IN FOOD PRODUCTS. Analytical data recalculated to parts per hundred of coloring matter present — Continued. GREEN TABLE NO. 435. LIGHT GREEN SF YELLOWISH. Determinations. Serial Nos. 57. 92. 168. 232. 0.04 0.10 1.88 1 04 Insoluble: Volatile .22 .00 .41 .00 .16 .65 .38 Nonvolatile .64 Total .22 .47 .81 1.02 Ether extract: Acetate .02 .05 .02 .05 .05 .05 .05 Acid 05 Total .07 .07 .10 .10 GREEN TABLE NO. 517. ERYTHROSIN. Determinations. Serial No. 216. Determinations. Serial No. 216. Common salt Ether extract: 11 Insoluble: 0.50 .22 Acid Total N onvolatile .11 Total .72 GREEN TABLE NO. 692. INDIGO SULPHO ACID. Determinations. Serial Nos. Determinations. Serial Nos. 90. 195. 249. 8. 31 90. 195. 249. Common salt 44.02 Ether extract: 0.02 .06 0.03 .15 02 1.25 . 59 1.23 .00 .77 .59 Acid .04 Total .08 .18 Total 1.84 1.23 1.36 MARKET QUALITY OF THE SEVEN PERMITTED COLORS. That the quality of the lots of the seven permitted colors of Food Inspection Decision No. 70 offered on the markets of the world after the issuance of that decision was no heller than that of the lots just reported, if as good, appears from the paper of E. G. Kohnstamm enti- tled, "Certified Food Colors: The Difficulties in the Way of their Manufacture/ 5 presented to the Seventh International Congrec Applied Chemistry held in London, May and dune, 1909, and pub- lished in abstract form. This abstract reads .-is follows: The coal-tar colors permitted under the food and drugs act of the United States ar< in Dumber, which must be is a high state of purity. The author slates thai of the colors on the markets of the vrorld, at the time oi testing, would meet these requirements. ANALYSIS- OF PERMITTED COLORS, 1907. 191 Of these seven colors, 189 samples, from every possible source, and representing all the leading manufacturers, are here reported on, and the results of their examina- tion reasonably establish the necessity of food color certification. Of these 189 sam- ples, the worst were offered for food coloring purposes and seemed to be so offered because unfit for any other purpose. Naphthol Yellow S. — Sixty-four samples examined, ranging in shade from clear bright yellow to a dirty brownish or green color. All contained Martius Yellow, some more than 1 per cent thereof; some contained as high as 2 per cent and even 3 per cent unconverted initial material or decomposition products. Forty-one contained exces- sive arsenic and 29 excessive heavy metals. Orange I. — Twenty-eight samples examined; all contained decomposition products varying from a slight amount to over 50 per cent; free r even suspicious, cither in chemical union • contaminations." The following page* Section XV) sbxw how closely this requirement has been met aa a result of quality control on the part of the Department of Agriculture. 192 J -TAB COLORS USED IN POOD PRODUCTS. XV. GUIDES IN DETEMUNIXG DEGREE OF PUEITY AND CLEANLINESS. In view of the absence of any statements in the literature defining the purity of the colors physiologically examined with such accuracy that another could obtain with reasonable certainty the same degree of cleanliness, it became necessary to devise some guide, no matter how empirical, in the setting up of standards, tentative or otherwise. As a first consideration it was held that, in view of the fact that all of the physiological work had been done with specimens of coal-tar colors of commercial purity, it would be reasonable to suppose that coal-tar colors produced in the purest form possible with present-day methods would certainly be as clean and as free from admixture as any of the commercial products subjected to physiological test. It •lonsidered unreasonable to expect that increasing the degree of purity of these substances could in any way increase any harmful property possessed by them. Certainly in the case of Xaphthol Yellows, where Martius Yellow is a usual contaminant, it can hardly be maintained that decreasing the amount of Martius Yellow would increase any harmful property which might reside in Xaphthol Yel- low S proper: in the case of Ponceau 3 R it could hardly be argued that any undisazotized base or decomposition products of diazo compounds tended to correct or counteract any harmful property that might reside in Ponceau 3 R proper; nor could it be maintained that Orange I free from uncombined alpha-naphthol was more harmful than Orange I, contaminated with alpha-naphthol, and so on through the list of the seven permitted colors. This point would not be raised had it not been pressed repeatedly by different persons as a serious objection to quality control and purity standards of the seven permitted colors of Food Inspection Decision No. 76. With this rule in mind, and referring to the tabulated results of the analyses of the 30 specimens of the seven permitted colors just given, the items in the analytical statements will each be separately 1. Common salt and ether extractive. — Common salt is a legitimate com- ponent of commercial brands of coal-tar colors in so far as I coal-tar colors are obtained by the so-called "salting out" process. The coal-tar colors are recovered from solution by the addition of common salt, which has the peculiar property of separating the coal- i olor from the water solution as an undissolved solid. The -^ar color so obtained will contain more or less salt, which, from a commercial manufacturing point of 1 be almost impossible to keep iron out to an extent which would bring the color within the pharmacopoeia! test. It has been shown that there were 1:5 colors (.n the market in 1907 which contained so little iron that, they failed to respond to the heavy metals test of the Pharmacopoeia in that respect; but here again the same criticism holds good as in the arsenic I hat there is no certainty that the amounts taken for the heavy metals tests were equivalent in each ca.se to 1 grain of actual coloring matter, and subsequent experience seems to confirm the correctness GUIDES IN DETERMINING PURITY. 197 of that criticism; therefore the rule for cleanliness of product, as just given, has been modified in that respect and to that extent. That it was only fair and reasonable to expect considerable improvement in the cleanliness and purity of food colors was made evident by an examination of two substances sold in the United States in large quantities for the purpose of making a very cheap red coal-tar coloring matter, which is used in many of the cheapest coloring operations, for paints, inks, etc. These substances are paranitranilin and betanaphthol. They were found in the United States market in such a condition of cleanliness and purity that had they been suitable for use in foods no objection could be raised against them on these scores. They complied with the requirements of the United States Pharmacopoeia with respect to freedom from arsenic and all heavy metals, inclusive of iron. The significance of this lies in the fact that all the raw materials entering into the manu- facture of paranitranilin and betanaphthol also enter into the manu- facture of the seven permitted colors of Food Inspection Decision No. 76, and that the only source of arsenic in food colors, and probably the only way in which iron could be introduced into them, is by way of the materials entering into the manufacture of paranitranilin and betanaphthol, and since it has been shown to be commercially possible to keep those bodies out of paranitranilin and betanaphthol, and since there is no occasion whatever for arsenic or iron or other heavy metal being present in any of the materials used in the manu- facture of the seven permitted colors of Food Inspection Decision Xo. 76, over and above the materials used in the manufacture of para- nitranilin and betanaphthol, there was every reason for believing that the seven permitted colors of Food Inspection Decision No. 70 could ultimately be manufactured and marketed in the same degree of cleanliness and purity that paranitranilin and betanaphthol are marketed; in oilier words, that food colors could be made as clean and as pure as paint colors, a, condition not existing in the food-color market of the United States in the summer of L907. The results of the control exercised by the Department of Agricul- ture over the quality of food colors, as compiled in the following section, fully justify such expectations, as well as the aim to make coal-tar colors when used for food purposes of the same high degree of cleanliness and purity as when they are to he used a- drugs or as they actually are when used for the man u fact \i re of paints and printer's inks; that is to say, that the coal-tar color used in a colored food should be as clean ;i- the coal-tar color used in making t he ink on t he label of Mich colored food, the very reverse of the .situation existing prior to the quality control now established 198 COAL-TAR COLORS L'SED IX FOOD PRODUCTS. XVI. ANALYSES OF CERTIFIED LOTS OF PERMITTED COLORS, 1909-10. TABULATION OF RESULTS. The analytical results obtained on 74 batches of certified colors, totaling upward of 32,000 pounds, are given in the following tables. The analytical results are the work of the New York Food and Drug Inspection Laboratory in the course of checking up the analytical results certified to in foundation and in supplementary certificates. The period covered by these examinations is approximately from July, 1909, to January 1, 1910. Naphthol Yellow S. [Figures calculated to 100 parts pure color and arranged in the order of their size. Figures on the same horizontal lines do not refer to the same samples; the dash line shows the location of the average of each column.] Insolubles. Ether extractives. Com- mon salt. Sul- phur. So- dium. Cal- cium. Number of batches. Total. Inor- ganic. Neu- tral. Alkali. Lcid. Total. 1 2 0.170 .130 .097 .090 .082 .079 0.100 .078 .050 .040 0.036 .026 .024 .023 .019 .018 0.016 .007 .006 .006 .005 .005 0.067 .065 .042 .038 .036 0.088 .073 .072 .064 .061 .057 .056 .054 4.06 2.26 1.68 1.41 1.30 9.23 8.99 8.98 8.90 8.89 13.27 13.21 0.13 .13 3 12.92 12.90 12.89 12.89 12.87 12.87 12.81 12.75 .05 4 .04 5 .040 .030 .032 .030 .010 .010 G .034 .031 .026 .025 .025 .020 .018 .63 .03 .48 .47 . 26 .24 .19 .19 S.86 8.85 8.71 8.70 8.62 7 .070 .000 .055 .050 .0.50 .030 .030 .016 .012 .011 .010 .007 .006 .004 .004 .000 .000 .000 .000 8 9 .048 .045 .045 .045 .044 10 11 . 12 13 Average Above Below .076 ti 7 .043 4 .017 6 6 .004 6 6 .035 5 7 .051 8 5 1.06 ."> 8 8.87 5 ti 12.93 2 8 .09 2 2 Ponceau .,/,'. [Figures calculated to KM) parts pure color and arranged in Hie order of their size. Figures in the same horizontal line do not refer to the same sample; Che dash line shows the location of the average of each column.] Number of batches. Insoluble. Fther extractives. Com- mon salt. So- dium sul- phate. Sul- phur. So- dium. (V.l- cium. Boiling point of cumi- din. Total. Inor- ganic. Neu- tral. Alkali. Acid. Total. 1 2 0.56 .43 .41 .27 0.29 .20 .17 0.166 .163 . L61 0.034 .033 . 032 0.019 .ills .ills .018 .017 .017 .016 .014 0.216 .218 . 201 :.. 73 6.87 0.06 .06 12.96 9.61 9.81 9.81 9. 28 0.18 .18 222-236 222-236 3 .00 .00 .(HI .no .00 .00 .on .14 .11 .12 .12 222-286 1 .12 .11 .11 .09 .07 .072 .OH .088 .009 .006 .002 .002 .(MM) .098 .073 .072 .068 .087 220-230 6 .17 .17 .It, .11 .10 4.12 3.73 3.71 3.(13 2. 63 12.80 12.80 L2.77 12.64 220 230 r, 9.16 '.i. L2 H.(i7 220-280 7 216-226 .266 4 B .16 B .III.", 8 t ! . 124 3 4.84 1 B ...: m 12. S3 4 t 9.21 B 4 .14 2 I 221-231 3 8 - ANALYSES OF CERTIFIED PERMITTED COLORS. 199 Orange I. [ Figures calculated to 100 parts pure color, and arranged in the order of their size. Figures in the same horizontal line do not refer to the same sample; the dash line shows the location of the average of each column.] Number of batches. Insoluble. Ether extractives. Com- mon salt. So- dium sul- phate. Sul- phur. So- dium. Cal- cium. Total. Inor- ganic. Neu- tral. Alkali. Add. Total. 1 0.39 .38 .28 .23 .23 0.18 .08 .08 0.224 .220 .210 .207 .181 .177 .174 0.093 .060 .043 .035 .035 0.08 .077 .067 .055 0.360 .309 .306 .300 ,2 .267 3.13 3.13 3.04 2.36 0.22 9.26 9.23 9.12 9.10 9.06 9.06 9.05 9.02 6.82 6.76 6. 62 0.29 2 .10 .24 3 .24 4 05 .04 .04 .03 .03 .02 5 .048 .039 .036 .029 .010 1.9S 1.91 1.90 1.41 1.05 1.01 6.54 6.48 6.32 6. 21 .05 6 .16 .15 .11 .10 .05 .020 .018 .015 .('11 .011 .04 7 . 246 .2-24 .175 .124 s .124 .107 .097 9 8.96 8. IS 10 Average Above Below .21 6 5 .06 3 .172 7 3 .034 5 5 .050 1 5 .258 6 4 2.09 4 6 .03 1 1 9.00 8 2 6.55 4 4 .18 4 2 Amaranth. [Figures calculated to 100 parts pure color and arranged in the order of their size. Figures on the same horizontal lines do not refer to the same samples; the dash line shows the location of the average of each column.] Insoluble. Ether extractives Com- mon salt. Sul- phur. Sodi- um. Calci- um. Number of batches. Total. Inor- ganic. Neu- tral. Alkali. Acid. Total. 1 0.43 . 30 . 23 .21 0.33 . 30 .19 .15 .10 .10 0.090 .088 .069 .017 0.030 .017 .009 .007 .017 .015 .014 .012 .012 .012 .101 .100 .060 4. is 4. lti 3.73 3.08 2.70 16.04 16.80 15.63 15.61 15.59 15.58 15.57 15.57 15.54 15.52 14.22 14.19 12. 10 o 20 2 a 17 4 11.47 11.43 11.43 11.42 11.38 11.33 11.32 11.81 1 1 28 11.27 11.27 11.21 11. is 11.1.-, 11 14 11.10 11 Ol .14 r, . 056 .053 .052 .051 .048 .012 .012 .010 .039 .0.-. 14 r, .13 .12 .12 .12 .11 .11 .11 . 11 • .070 .07 .035 .034 .034 .034 .027 .026 .021 .012 .005 .005 .005 .(Mi:, .mil .(HM1 13 : .06 .06 .04 .04 .03 .08 .02 .02 13 .010 .007 .005 .003 12 12 10 2. 62 2.21 2. 12 2. 12 1.U4 1.91 L01 12 n 11 14 15.41 15.41 15.41 IV lo 15. 12 .07 .00 15 16 17 20. 21 .148 11 .006 7 12 4 r, 11 13 B .012 7 11 200 COAL-TAR COLORS USED TX FOOD PRODUCTS. Light Green SF Yellowish. [Figures calculated to 100 parts pure color and arranged in the order of their size. Figures in the same horizontal line do not refer to the same sample; the dash line shows the location of the average of each column.] Number of batches. Insoluble. Ether extractives. Com- mon salt. Sodi- um sul- phate. Sul- phur. Sodi- um. Calci- um. Total. Inor- ganic. Neu- tral. Alkali Acid. Total. 1 0.07 0.034 .02 0.058 0.011 .007 0.011 .009 0.076 4.15 0.00 .00 .00 .00 .00 .00 12.69 12.27 8.03 7.99 7.99 7.94 7.73 4.52 2 .041 .034 .033 .054 .053 .042 .64 .515 .43 .37 .10 3 4 . 03 .02 .02 .01 .01 .01 .01 .01 .00b .000 .008 .007 12.07 11.88 11.88 .09 .06 .OS 1.56 Aver . Below .033 2 4 .017 2 3 .041 3 .006 2 2 .008 2 2 . 056 1 3 1.221 4 .00 12.16 2 S 6.87 5 1 . 96 1 Erythrosin. [Figures calculated to 100 parts pure color and arranged in the order of their size. Figures in the same hoii- zontal line do not refer to the same sample; the dash line shows the location o the average of each column.] Insoluble. Ether extractives. Com- mon salt. Sodi- um sul- phate. Sodi- um. Number of batches. Total. Inor- ganic. Neu- tral. Alkali. Acid. Total. Iodin. 1 0.13 .09 0.09 .04 0.039 .039 . 037 0.017 .009 .oos 0.051 ; .048 047 1.13 1.00 .48 0.00 .00 .00 .00 .00 .00 .00 .00 .00 5.42 5.35 5.33 5.29 5.15 5.11 2 3 . 059 .05 .04 .04 .04 .03 .024 .01 .01 .01 .01 56. 40 4 .034 .033 .033 .027 . 002 .002 .000 037 .23 .09 .09 .09 .09 .036 56. 07 035 035 55.98 6 55.97 7 .027 3.85 55.94 55.90 g .059 2 6 .028 2 5 . 035 3 4 . 006 3 3 040 3 4 . 360 3 6 .00 5.07 6 1 56. 23 3 6 1 Indigo disulpho acid. [Figures calculated to loo parts pure color and arranged in the order of their size. Figures on the same hori- zontal lines do not refer to the sumo samples; the dash line shows the location of the average of each column.] Number <>f lies. Luble. Ether i Com- mon salt. Bodi- um- sul- phate. Sul- phur. Sodi- um. Calci- um. Total. Inor- ganic. Neu- tral. Alkali. Acid. 'total. 1 . . . . o. 50 . It, 0.28 .21 O. 1 10 .Old 0.320 .217 6.77 16. 12 10.22 18. 15 18, 12 lo. 18 9.90 9.64 0. 45 .026 .021 .00, .13 :< . 102 .088 .07.", .015 .018 .113 . 1 II 4.79 .110 .00 .00 .43 . 32 2. 89 1.85 . to . i ).', . 22 . 16 13. I.", 9. 13 Y 5 1 i, A\ci Bekw . :< . 235 1 2 3 2 8 .011 i .189 •> 8 3. 1 2 3 1 I 9.49 1 2 . L9 1 3 ANALYSES OF CERTIFIED PERMITTED COLORS. 201 COMPARISON WITH ANALYSES MADE LN 1907. A comparison, as far as possible, of these figures with the corre- sponding data previously given for samples on the market in the summer of 1907 is made in the following table. The figures in this new table show the value of the fraction obtained by dividing the old average figure by its corresponding new average figure — that is, they show how many times greater the old average figures are than the corresponding new average figures. Italics indicate those items in which there has been a retrogression hi the new figures as com- pared with the old; all the other figures represent an advance. It will be noted that comparisons are not made for Ponceau 3R and Erythrosin. Tins comparison was omitted because there was but one sample of Ponceau 3R examined and a partial analysis of one sample of Erythrosin in the old work. Ratio of average figures of 1907 to those for the certified colors, 1909. Salt. Insoluble. Ether extr u Color. Total. Vola.ile. *E* Nonacid. Acid. Total. Naphthol Yellow S 4.88 3.54 15.07 .66 8.05 4.44 9.07 11.26 31.00 3.29 9.02 0.67 7.50 °l-89 15.97 Orange 20.00 21.40 29.00 5.4 3.00 7.71 17.00 1 . 57 4.07 .16 10.77 5.00 Green Blue These retrogressions are: 1. Salt in the Green. 2. Nonvolatile insolubles in the Yellow. 3. Nonacid ether extract in the Green. 4. Nonacid ether extract in the Blue. 5. Total ether extract in the Blue. With respect to the first of these retrogressions there is this to be said: One lot of Green had apparently been purified by precipitation with salt, since it was in every other respect of a high quality that is, it was free from arsenic and heavy metals within the pharmacop and its ether extractives were very satisfactory. The other lots rreen examined had apparently not been made in this manner. With respect to the second retrogression, the probable explanation is that some of the lots were made during a period when the municipal water supply was excessively hard, as before explained. The remain- ing three retrogressions are probably due to the fact that the methods of analysis used were not so accurate as the methods later employed and hereinafter described (pp. 223 It will be noted that in three of the six batches of Indigo disulpho acid reported there is no sodium Bulphate; whereas in the other three lies the amount of this substance is as high as 16.12. The for this LS that in the early Btages Of the work the results obtained 202 COAL-TAK COLORS USED IN FOOD PRODUCTS. by the manufacturers pointed in the direction of the impossibility of getting rid of all the sodium sulphate; but later and more extended work showed tins conclusion to be an error. There is therefore no good reason for permitting sodium sulphate as a contaminant in this or any other of the seven products. CONFORMITY OF ANALYTICAL DATA WITH THEORETICAL COM- POSITION. The seven tables following, likewise based upon the 74 Govern- ment analyses, show the conformity or nonconformity, as the case may be, of these samples with their theoretical composition. The second and third columns show the percentage based on theory of the contained sulphur and sodium in the case of all colors except Erythrosin, where the figures for iodin are substituted for those of sulphur. The fourth column shows the actual ratio of sodium to sul- phur or iodin, as the case may be, and the fifth column gives the percentage of the figure of column 4 based on the theoretical value winch is given in parentheses at the head of colunms 4 and 5. The summary gives the maximum, minimum, and average of each column. Taken as a whole the figures are fairly satisfactory and show a rea- sonable conformity of the actual product to theoretical requirements, although there seems to be considerable room for improvement, which no doubt can be achieved in time. 1'iircntages based on theoretical composition. NAPHTHOL YELLOW B. Number of Sulphur. Sodium. Na s =(1.4379). Number of batches. Sulphur. Sodium. Na -g =(1.4379). batches. Value. Per cent. Value. Percent. l Per cent. 103. 120 90.312 97. 205 Per cent. 103. 10 9'.). 068 100. (MM) 1.4:577 1.4791 1.4792 1.4707 99.983 102.860 102.870 102. 270 10 11 12 13 Per cent. 98.880 98.658 KM). 330 100. 100 Per a ui. 101.06 101.22 100.15 100.00 1.4578 1.4631 1. 1353 101.36 2 101.74 3 4 97. 642 'W.VKI .-, Average. . Min 99. 101 103.120 90.312 102.212 1 03. .->00 1.4.j7S l. L792 101. S7S 7 98. 992 99.331 1(H). HI) 8 100.15 103.50 1.4498 100. SO 102. IS 99.82G 9 PONCE N urilxT of Sulphur. Sodium. 0.719..:,). Number of Sulphur. Sodium. Value. Per cent. Value. Percent 1 Pert i nt. J'i r a nl. 93.617 9a i7i 8 9 Avenge. lias .Mm ... l'ir cent. Per ct ui. 101.920 L01.700 0.74296 103.33 102.48 100. 110 .70518 .71892 98.662 99.918 lol. 16 4 99. 266 11MI. 10 97. AM 98. 73S 101.920 98.617 .Tints 99. 144 1 f. ANALYSES OF CERTIFIED PERMITTED COLORS. 203 Percentages based on theoretical composition — Continued. ORANGE I. Number of Sulphur. Sodium. Na s =(0.71905). Number of batches. Sulphur. Sodium. Na .71905). batches. Value. Per cent. Value. Percent. 1 2 z. '.'.'.'... '. 4 s 88.972 100.76 101.09 99. 135 99. 135 97.810 98.799 100.44 103. 72 102.53 0. 81226 .73002 112.97 102. 77 101.53 8 9 10 Average . Max Min 98. 471 99.563 99.342 85.901 99. 241 98. 330 0.70066 .71710 .71208 97.454 99.740 99.042 98.308 101.09 88.972 99.376 103. 72 95.901 .72957 .81226 .69309 101.475 o 7 94. 234 100.610 .69309 . 73243 96. 391 101.90 112.97 AMARANTH. , 96. 795 105. 76 0. 78519 2 98. 176 99. 037 .72488 3 94. 975 124.29 .94048 4 94. 975 124.04 . 93849 o 100.750 96.240 .68637 b 99.246 99.911 . 72341 7 97. 174 100.260 . 72454 8 96.929 99. 827 .74011 9 97. 926 97. 465 .71820 10 96. 975 97. 990 . 72744 11 96. 733 97. 389 .72338 12 97. 487 98.514 .72615 13 96.798 97. 028 . 72030 109.21 100.82 130.80 130.53 95. 470 100.62 100.78 102.94 99. 476 101.18 100.62 100.99 100.18 Average . Max Min 97.800 97. 499 97. 614 97.499 97.863 98. 052 97.800 97. 735 97.467 100.750 94. 975 98. 863 99.911 98.950 99.476 98.689 98. 393 98.514 0. 72640 .73646 .73325 . 72465 .72106 101.347 124. 29 96.240 . 74707 . 9404S loi. as 102. 43 101.32 101.98 loo. 78 100.05 100. 68 99.914 103.90 130.80 95. 470 LIGHT GREEN BF YELLOWISH. 1 102.85 6 104.50 95. 547 0.65783 2 98.148 93.018 98. 020 96.148 3 4 5 109.87 106.23 102.35 0.60911 .65358 90.911 93. 540 • Average. { Min 105. 26 109.87 95. 498 96.029 93.018 .6482/ ?67256 .60911 90.167 ERYTHROSIN. Number of Iodin. Sodium. 1 2 Number of batches. Iodin. Sodium. 090855). batches. Valt if. Per cent. 1 2 a 4 5 Per cent. 98.915 97.061 97.780 97.209 97.035 Per cent. ft. 473 98. 277 1 .000604 75. so;, 104.00 1. 9 \ \« rage. Min 96. 879 102.09 .005741 97. 952 73. 178 .090204 99.146 <; 101.71 204 COAL-TAR COLORS USED IX FOOD PRODUCTS. Percentages based on theoretical composition — Continued. INDIGO DISULPHO ACID. Number of Sulphur. Sodium. ^ :1 = (0.71905). Number of Sulphur. Sodium. N * a = (0.71905). S batches. Value. Per cent. batches. Value. Percei t. 1 2 3 4 5 Per cent. 97. 670 95.636 97.600 96. 148 97. 382 Per cent. 103. 330 98.279 103. 760 91.926 106.560 0. 71482 .69430 .71832 .64598 . 73934 99.423 95.669 99.910 89.848 102. 830 6 Average . Max Min Per cent. 97. 818 Per cent. 109.040 . 75315 104. 750 97.042 97. 818 95.630 102. 149 109. 040 91.926 .71099 .75315 .64598 98.738 104. 750 89.848 ARSENIC DETERMINATIONS ON 86 BATCHES. During the course of tliis work it became necessary to determine the exact amount of arsenic contained in 86 of the various batches of certified colors; for this purpose the arsenic method of Seeker and Smith (see p. 212) was devised. These results are expressed in the number of parts of color containing one part of metallic arsenic (As). The numbers in parentheses indicate the number of specimens of the quality indicated. Xaphthol Yellow S (17 specimens).— 770,000, 833,000, 1,250,000 (2), 1,428,000 (2), 1,666,000 (3), 2,000,000 (6), 3,333,000 (2). Ponceau (15 specimens).— 588,000, 625,000, 667,000, 714,000 (2), 769,000 (2), 1,000,000 (2), 1,111,000, 1,250,000, 1,666,000 (2), 2,000,000, 2,500,000. Orange (7 specimens).— 200,000, 250,000, 588,000, 1,000,000, 1,111,000, 1,429,000, 5,000,000. Amaranth (27 specimens).— 909,000, 1,000,000, 1,111,000, 1,250,000 (2), 1,438,000 (3), 2,000,000(4), 2,500,000 (4), 3,333,000 (9), 5,000,000 (2). Green (8 specimens). —166,000, 200,000, 370,000 (2), 500,000, 666,000, 770,000, 833,000. Eryihrosia (7 specimens).— 2,000,000(2), 5,000,000 (2), 10,000,000 (2), 20,000,000 (1). Blue (5 specimens).— 285,000, 666,000, 1,428,000, 3,333,000 (2). The United States Pharmacopoeia test for arsenic is sensitive to 0.005 mg of arsenious oxid (As 2 3 ) which on a sample containing 2 grams of actual or real color would amount to one pari of arsenious oxid in 400,000 of color; calculated to metallic arsenic, the basis employed in the foregoing, this would mean 528,000 parts of color for each one part of metallic arsenic. There are, therefore, in the foregoing 86 lots of certified colors hich did not comply with thai requirement, namely: Orange (2 specimens).— 200,000 and 250,000. Green (5 specimens).— 166,000; 200,000; 370,000 (2); 500,000. Blue (l specimen). — 285,000. The reason for this discrepancy is thai in the preparation of the samples for analysis by the United Stales Pharmacopoeia test a loss of arsenic ensued, which is avoided in the Seeker-Smith method now employed in making these determinations. Had the existence ANALYSES OF CERTIFIED PERMITTED COLORS. 205 of this discrepancy been proven at the time the first analysis was made certification would have been denied to the eight lots above mentioned. These results further show the position taken early in the work by several of the manufacturers — that a requirement of not more than 1 part of metallic arsenic in not less than 264,000 parts of coal-tar color could not be complied with on a commercial scale to be untenable; on tins basis only 4 out of the 86 lots examined would have been excluded. Further, the position of some manu- facturers and dealers that the arsenic requirement ought, for prac- tical manufacturing reasons, not to be more rigorous than one part of metallic arsenic in 26,400 parts of color, or 1 part of arsenious oxid (As 2 3 ) per 20,000 parts of color, is not borne out by the data. SUGGESTED REQUIREMENTS FOR CERTIFIED COLORS. Although the material embodied in this report gives a very good idea of the composition and quality of substantially 30 different lots of permitted colors prior to the issuance of Food Inspection Decisions Nos. 76 and 77, and of 74 lots of certified colors, yet these data are hardly sufficient to furnish a basis for standards with winch each color specimen must comply in detail. The fitness or unsuit- ability of any lot has been determined by the examination of the analytical data obtained thereon in the Food and Drug Inspection Laboratory at New York; such examination has been applied to the particular color under investigation with respect to its general relationship to the results theretofore achieved. If in some minor quality, as, for example, freedom from salt, the sample was not up to what had been previously accomplished, but in a major quality, as for example, ether extractive, it was equal to or better than what had been previously accomplished, and the pharmacopoeia] tests for arsenic and heavy metals were satisfied with the exception of iron, and the amount of iron was within die limit previously stated, 0.005 per cent, and the other factors showed a fairly close conformity, such a defect as its salt content would not act as a bar to the pass- ing of the lot; however, no matter how good a color might be in respect to such determinations a- ether extractive, if it failed to comply with the United States Pharmacopoeia requirement for arsenic or for heavy metals it was not accepted. These arsenic results have been tabulated to show the distribution of arsenic content (As,O t )j the numbers at the top are the Green Table numbers; the numbers in the body of the table indicate the number of specimens of the arsenic content sti ted in the first column ; the last column shows ihe totals of all colors of the arsenic content (As_(),,) corresponding thereto. 206 COAL-TAE COLORS USED IK FOOD PRODUCTS. Arsenic content of S6 lots of certified colors. Arsenic content. Green Table numbers. Total. 4 56 85 107 436 517 692 1 part in— 1 126,200 i 151.000 1151,500 i 189, 300 1215,900 i 280, 300 i 378, 700 44."). 400 473,500 504,600 505, 300 540, 900 582, 500 583,300 631,000 688, 700 757,500 841,600 946,900 1,081,000 1,082,000 1,089,000 1.262.000 1,515.000 1,893,000 2,524.000 3,787,000 7.157,000 15, 150, 000 il il 1 1 1 1 1 2 1 2 1 1 2 2 2 2 1 4 3 5 3 3 5 13 5 13 5 2 1 1 1 il l 2 i 1 1 1 1 1 1 ...... 1 2 2 1 1 1 1 1 2 2 2 1 1 1 1 1 1 3 3 6 2 1 4 4 9 2 2 2 1 2 2 1 17 15 7 27 8 7 5 86 i The certification of these lots was due to an unknown source of error in the analytical method; the analyses made at that time (July, 1909), showed less than 1 part of AS2O3 per 400,000; see also page 204. It is therefore somewhat premature to attempt to define rigidly the requirements for composition and purity for colors until a suffi- cient number of analyses is available to permit a hard-and-fast line to be drawn for each item as required for each color. Until that time the decision as to whether or not a certain color shall be cer- tified must rest with the Department of Agriculture. However, the following requirements are tentatively suggested as being com- mercially practicable. It should be clearly understood that the ten- tative requirements here stated are based on the results of actual control, and are not any more searching or numerous than arc the requirements for many if not most of the coal-tar dyes or their com- ponent parts in the industrial arts, particularly for the various kinds of paint, varnish, and ink making. While it may be that some of the tentative requirements herein defined necessitate the expenditure of considerable work and time, yet that is also true of some of the commercial requirements, since manufacturers of cheap paints, varnishes, inks, and the like, find it wise and necessary carefully to control i\w quality of the coal-tar dyes or their component parts which they use, it can not be less wise or accessary to extend the same kind of quality control to those coal-tar dyes intended and sold ANALYSES OF CERTIFIED PERMITTED COLORS. 207 for human consumption as food. This stand is fully justified by the quality of the coal-tar dyes offered as food colors on the United States markets as described in the foregoing pages. That such control is not only practical, but practicable, is fully proved by the fact that more than 20.5 tons (41,000 pounds) of coal-tar food dyes have been so controlled, examined, and certified under the food inspection deci- sions hereinbefore mentioned. The figures are expressed in parts per hundred of actual color contained in the dye and not in parts per hundred of the total substance. The numbers at the head of the columns are the num- bers in the Green Tables. Tentative limits of composition suggested for permitted colon. (Parts per hundred of actual color.; Determinations. Insolubles: Total 0. 070 Nonvolatile on ignition Ether extractives: Neutral Alkaline Acid Common salt Green Tables numbers. ■ 107 43.5 - 692 0.070 0.270 0.250 0. 130 0.030 0.000 0. 4.50 .040 .150 . 050 .050 .010 .020 .250 .017 .090 . 150 .030 .040 .035 .100 .004 .015 .03.5 . 00.5 .006 .002 .030 .060 .017 .050 .010 .008 . 150 .000 5.000 2.000 2.500 .600 .400 3.000 In addition to these there are the following requirements applicable to all colon : 1. The absence of admixed dye must be convincingly demonstrated by suitable test. 2. Arsenic. — Test 17 of the United States Pharmacopoeia, 1900, applied to so much of the specimen as represents 3.5 grama of actual dye must give a negative response. Such negative result must be reenforeed by a check test identical with the test on the dye with the addition of 0.005 mg of arsenic (\s,0 3 ) to the dye prior to treatment, and this check test must produce a positive result for the presence of arsenic. Igni- tions in the preparation of ihe material for the test must be made in porcelain. Any other mode of testing which is demonstrated to be capable of detecting 0.005 in.-.! Ided arsenic (AftjOj) in so much of the specimen as represents 3.5 grams of actual color will, of course, 1"' accepted. This quantity, how e ve r , is only tolerated tenta- tively pending further investigations relative to the complete, or practically com- plete, elimination of arsenic from foods, especially those which are largely used by children, such as candies. .".. linn ij in, tuis. Test l-i of the United States Pharmacopoeia as revised May l, L90T, using so much of the specimen as represents one part of actual dye must negative response for all metals except iron, which may be present in amounts I of 0.006 per cent of iron based on t he d\ e actually pr e s en t in the specimen. 4. None of the d; 1 for certification shall contain any dauber's salt or sodium sulphate in any farm, nor shall they contain any added BUgar, dextrin, or other loader, tiller, or reducer for any purpose whatsoever, and Convincing proof Of the . any or all of them must DO Submit 208 COAL-TAR COLORS USED IN FOOD PRODUCTS. 5. The ether extractives are to be made successively upon water solutions of the dye, first neutral, then made alkaline with caustic soda, and then made acid with hydrochloric acid, using washed or sodium, dried ether. 6. The sulphur content of the sulphur-containing dyes must agree substantially with the theoretical: likewise the sulphated ash figures of all must agree substantially with the theoretical; variations between these two sets of figures, as long as they are stent with each other, will not be reason for rejection. 7. In the case of No. 4 proof must be submitted showing that the specimen is of the sodium or potassium variety, and if it is a mixture of these two varieties the proportion of each present in the mixture must be stated. 8. In the case of Xo. 5G the crude cumidin employed may have a boiling point of from 220° to 230° C, and may be liquid or solid; the absence of any compound of S or G salt must be convincingly shown. 9. In the case of No. 85 convincing proof must be submitted that beta-naphthol orange if present at all is present in an amount not in excess of 5 per cent of the coal- tar dye present. 10. In the case of Xo. 107 the absence of any compound of S or G salt must be convincingly shown. 11. In the case of Xo. 435 the product should be free from calcium ; convincing proof of absence of Xo. 434 must be submitted. 12. In the case of Xo. 517 the actual dye must contain not less than 56 per cent of iodin (sodium basis) and must not contain any other halogen; the kind and amount of metallic base, whether sodium, potassium, or the like, must be shown. 13. In the case of No. 002 the absence of indigo monosulphonic acid and of nonsul- phonated indigo must be convincingly shown. 14. Each foundation certificate must be tiled in duplicate, but need not be executed in duplicate and must contain a summarized or tabulated statement of all the quan- titive results contained in the certificate, also a tabulation or summary of the quali- tative tests made, together with the results of such tests, all stated on one sheet, so that the certificate will bear within itself its own summary and conclusions. 15. The fundamental analytical data must be given with such fullness as to permit efficient checking of the calculations, and the arithmetical operations performed should be indicated wherever needful to avoid confusion, or to facilitate the work of the checking chemist, or make the meaning of the certificate more plain. There are freely quoted in the United States market two substances, paranitranilin and betanaphthol, which are subject to much com- petition, the prices for which, wholesale, are not far from 25 cents and 9 cents, respectively. A specimen of each lias been examined, and, as before stated, they have both been found to be of such purity, with respect to arsenic, heavy metals, and general cleanliness. that had they been capable of use in food products, no objection against their use on this score could reasonably be raised; certainly no such objection could be successfully maintained. The following table discloses the chemicals entering into the manu- facture of paranitranilin and of betanaphthol, as well as of each of the seven permitted colors; the ingredients numbered 1 to 7 are used in the manufacture of these two substances, as well as in the seven permitted colors, as indicated by the "x" entries; ingredients 8 to 20 are used only in the manufacture of the seven permitted colors, and not in paranitranilin and betanaphthol. ANALYSES OF CERTIFIED PERMITTED COLORS. 209 Comparison of chemicals entering into the composition of the seven permitted colors and of paranitranilin and betanaphthol. Determinations. 4. Xap- thol Yellow S. 56. Pon- ceau 3R. 85. Orange 107. Ama- ranth. 435. Light Green. 517. Ery- thro- sin. 692. Indigo disul- pho acid. Parani- trani- line. ■& Q. O X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X I X 2. Benzol 3. Sulphuric acid ' x X 5. Metallic iron ' 6. Acetic acid X X X 7. Caustic soda or potash ... x X Carbonate of soda Common salt Wood or methyl alcohol. Lime Hydrochloric acid Sodium nitrite Ethyl chlorid Peroxid of lead or of manganese Phosgene Iodin Metallic mercury Ferric chlorid Chlorin 11 Id Id With respect to the arsenic content of the finished product, it can be asserted, without any fear whatever of successful contradiction, that the arsenic finds its way into the goods by way of the sulphuric acid which is used in the manufacture of all the permitted colors, as well as in the manufacture of Paranitranilin and Betanaphthol. If arsenic from this source can be kept out of these substances it can also be excluded from the permitted colors. Turning now to those ingredients below the parallel lines, the only means of introducing arsenic would be through the hydrochloric acid used, which in turn derives its arsenic from the sulphuric acid used in its manufacture, and since arsenic free-sulphuric acid can be used in the first stages of producing the seven permitted colors and tho paranitranilin and betanaphthol, itcan also be employed in the manu- facture of the hydrochloric acid used in the subsequent stages of manufacture of the seven permitted colors. Therefore, it seems unreasonable to permit a higher arsenic content in food colors than in paranitranilin and betanaphthol, which are sold in open com- petition, and which are used for the production of tho very cheapest colored cloths, inks, and paints. With reference to the content of iron, and other heavy metals which satisfied the pharmacopuial tests in the ease of paranitranilin and betanaphthol, these materials probably enter the product from tho vessels in which the manufacturing operations are performed. The same hind of vessel used in the manufacture of paranitranilin •1 - -Bull. 147—12 14 210 COAL-TAR COLORS USED IN FOOD PRODUCTS. and betanaphthol is used in making the seven permitted colors, and since the heavy metals can be and are kept out of these two substances to the extent required by the Pharmacopoeia, there seems to be no good reason why they should not be kept out of the seven permitted colors when made for use in food products. Since all these defects in the seven permitted colors can be obviated in their first stage of manufacture (as is shown in the case of para- nitranilin and betanaphthol, and in some actual commercial samples of 1907, see sections XIV and XV), they are commercially avoidable at subsequent stages of manufacture, and there is no good reason why they should be then introduced. XVII. METHODS OF ANALYSIS USED IN TESTING COLORS FOR CERTIFICATION. INTRODUCTION. The exact analytical methods developed and tested in the New York Food and Drug Inspection Laboratory and used in obtaining the analytical data contained in the preceding chapter are described in the following pages. These descriptions were written by A. F. Seeker, of that laboratory, under whose immediate supervision all the laboratory work was done, and they represent a great deal of work, extending over more than three years. The methods are submitted in the hope and expectation that in their wider application by a larger number of chemists any defects in the methods or conclusions drawn from the results will be detected and rectified. Experience in the New York laboratory has shown that even different chemists of varying degrees of experience in this particular line of work obtain concordant duplicates after a comparatively short laboratory acquaintance with these methods. The identity of these colors and their freedom from foreign dyes is shown by the close agreement of their elements, as determined by analysis, with their theoretical composition; their behavoir toward reagents as given by standard works on dyes; their distribution between the layers when neutral, alkaline, and acid aqueous solutions are shaken with various immiscible solvents; uniformity of shade in the spots produced by particles of the dry color blown over the surface of wet filter paper, or water, and over concentrated sulphuric acid; uniformity of shade produced by a 0.5 per cent dyeing on wool under standard conditions, with similar dyeings from fractions produced by partial precipitation, by partial salting out, by fractional crystalliza- tion, and by extraction with alcohol or some liquid in which the pure color is not very soluble; and the behavior of the dye in acid and alkaline solution toward cotton. Which of these tests are Deeded to prove conclusively the identity of certain dyes or to establish their METHODS OF ANALYSIS FOE CERTIFIED COLORS. 211 absence, is a matter that varies so much from case to case that it must be left largely to the individual judgment to decide on the best com- bination of tests, and for that reason such combinations are not here offered. The general methods and procedures just outlined have, however, when properly combined, led to satisfactory results. In order to compare the results of the color analyses on the same basis, the actual figures obtained in the various determinations besides being reported as found, are also recalculated on a basis of 100 parts of actual color, i. e., the sum of the percentages of material which is not coloring matter, such as moisture, total insoluble matter, sodium chlorid, sodium sulphate, and ether extractives is deducted from 100 and the percentages of the various constituents found divided by the difference, the quotient being then multiplied by 100. The percentage of sodium in actual dye is calculated from the sul- phated ash. The methods of analysis of the seven permitted colors of Food Inspection Decision No. 76 are given in the following order: 1. Naphthol Yellow S. 2. Ponceau 3R. 3. Orange I. 4. Amaranth. 5. Light Green S F Yellowish. 6. Erythrosin. 7. Indigo Disulphoacid. NAPHTHOL YELLOW S. MOISTURE. Dry from 1 to 2 grams of the finely powdered dye in an air oven at 120° to 125° C. to constant weight. TOTAL INSOLUBLE MATTER. Dissolve 5 grams of color in 200 cc of hot distilled water, filter through a tared gooch, wash till the washings run through colorless, dry the insoluble residue at 105° C, and weigh. NONVOLATILE OR INORGANIC INSOLUBLE MATTER. Ignite the gooch containing the insoluble matter of the last deter- mination at a low red heat, cool, and weigh. SODIUM CHLORID. Dissolve 3 to 5 grams of dye in 200 cc of water, acidify the solution with nitric acid, and precipitate the chloric as silver chlorid. The latter is separated, unshed, ignited, and weighed in a tared gooch crucible in the usual w ay. SODIUM SI i rn ate. Dissolve i gram of dye in about 100 cc of water in a 200-ce gradu- ated flask, add 10 cc of a 20 per cent solution of potassium chlorid, shake, the mixture well, make up t<> mark with water, shake agai^ 212 COAL-TAR COLORS USED IN FOOD PRODUCTS. and then filter through dry paper. Treat 100 cc of the filtrate (repre- senting 0.5 gram of dye) with 5 cc of 10 per cent barium chlorid solu- tion without acidifying, and allow to stand overnight. If a precipi- tate forms, it is filtered, washed, ignited, and weighed in the usual way. HEAVY METALS. Treat the sulphated ash from 1 gram of the sample with 20 cc of water, digest with 10 cc of 10 per cent hydrochloric acid till solution is complete, place 3 cc of the mixture in a test tube, add 10 cc of freshly prepared hydrogen sulphid test solution (U. S. P.), shake the mixture, warm to 50° C, stopper, and allow to stand in a warm place (at about 35° C.) for half an hour. Run a blank test at the same time, with the same amount of hydrogen sulphid solution, using water instead of the solution containing the color ash. No turbidity other than that sometimes produced by slight separation of sulphur should appear in this test. Both tubes are then made slightly alkaline with ammonium hydroxid, and no precipitate should be produced, although a slight coloration, due to the presence of a small amount of iron, sometimes occurs. If this coloration is very marked the amount of iron should be determined. This is done by digesting the sulphated ash from a weighed amount of the sample with hydrochloric acid until all of the iron has gone into solution. The solution is filtered, and the filtrate poured into an excess of hot, pure, freshly prepared sodium hydroxid (by sodium) solution in a platinum dish. The precipitate is washed, dissolved in dilute hydrochloric acid, and again precipitated with ammonium hydroxid. The last precipitate is washed, ignited, and weighed in the usual manner. ARSENIC (SEEKER AND SMITIl's METHOD). Dissolve 10 grams of the dye in 200 cc of water, heating to insure complete solution of the color, add about 10 cc of strong bromin water to convert any arsenite to arsenate. Make the mixture alkaline with a few cubic centimeters of strong ammonium hydroxid. Twenty cubic centimeters of a sodium phosphate solution containing 100 grams of crystallized sodium phosphate per liter are next added from a pipette, after which magnesia mixture (containing 55 grams of hydrated magnesium chlorid, 55 grams of ammonium chlorid, and 88 CO of ammonium hydroxid, specific gravity 0.9 per liter) is added from a burette, stirring vigorously. The amount of mag- nesia added should be in slight excess of that necessary completely to precipitate the phosphate and should previously he ascertained hy blank experiment. Then add 10 CC of ;inmionium hydroxid (specific gravity 0.90), and allow the whole to stand for at least, three hours; separate the precipitate hy tiltration and wash it free, or nearly so, of METHODS OF ANALYSIS FOR CERTIFIED COLORS. 213 ,GLASS ruB£ 2.5 mn K dye with ammonium hydroxid containing one-tenth its volume of ammonia (specific gravity 0.90). Dissolve the precipitate from the paper with 1 : 1 nitric acid, the washings being collected in a large porcelain crucible; add 5 cc of sulphuric acid to the contents of the crucible, and evaporate the whole almost to dryness. It is not nec- essary that the solution should be colorless at this point, a brown colored solution giving equally accurate results. Add 20 cc of water to the residue in the crucible and then 10 cc of a saturated solution of sulphur dioxid; evapo- rate the solution to a sirupy consistency to remove the sulphur dioxid, and then take up in 20 cc of water and place in a 30-cc evolution bottle, add 5 cc of concentrated sulphuric acid, and determine the arsenic in the form of appa- ratus (fig. 1) used by Bishop in his modification of the Gutzeit test, the stains obtained being compared with those given by known amounts of arsenic. The apparatus used by Bishop consists of the following parts: A 30-cc salt mouth evo- lution bottle into which is fitted a one-hole rubber stopper carrying a glass tube 6.5 cm long with an internal diameter of 1 cm, this tube in turn being provided with a one-hole rubber stopper fitted with another tube of the same diameter and 5.5 cm long, the diameters of both tubes being constricted at the points at which they are inserted in the rubber stoppers. A third glass tube 15 cm long, having an in- ternal diameter of 2.5 mm, is fitted into the second at its upper end by means of a rubber stopper. The first tube contains a strip of filter paper which has been saturated with a 5 per cent solution of lead acetate and dried. The second tube contains a loosely packed plug of cotton-wool freshly moistened with a 1 percent r „ : . i._Appmtaa for the solution of Lead acetate. Into the topmost tube aetanniMttaiofiiwoto. is inserted the strip of sensitized paper to receive the anenic stain. The arsin is generated by introducing into the evolution bottle six pieces of Kahlbaum's stick zinc (arsenic-free for forensic purposes) weighing in all about s grams, and t<> assist in an active and constant evolution of L r ;:s a disk of platinum i- also [.laced in each l>«>nl<> to form an elec- trolytic couple. 'I'll,- evolution of gas IS allowed to proreed for one hour. The stains are produced on .strips of hard pressed w hite paper (2 nun wide and 120 mm long) that has been sensitised by being H7TH /%*>S.*C£T>4TT: ^-SO/Pf /crr» f>a. *c£-n*nr fww OZ BOT-7ZS 214 COAL-TAR COLORS USED IN FOOD PRODUCTS. dipped in a 5 per cent alcoholic solution of mercuric chlorid and then dried. (Note. — It has recently been found that mere uric brornid yields stains that are more evenly distributed and also produces standards that are incomparably more permanent.) For purposes of com- parison it is better not to develop the strips stained by the arsin, though some prefer to dip the stains in ammonium hydroxid, which causes them to become black. A blank is run with each set of determinations. ETHER EXTRACTIVES. Dissolve 10 grams of color in 150 cc of water and extract in a separa- tory funnel with ether that has been washed with water (using three 150 cc portions of water for each liter of ether). Extract the color solution with two 100 cc portions of this ether, shaking thoroughly for one minute, and wash the combined ether extract successively with 35, 20, and 10 cc of water made alkaline or acid, as the case requires, with 1 cc of tenth normal alkali or acid per 100 cc of water. Decant the ether from the mouth of the separatory and rinse the funnel once with 5 cc ether. The color solution is first extracted neutral, the extracted solution being then rendered alkaline with 2 cc of a 10 per cent solution of caustic soda and again extracted with two 100-cc portions of ether. In acidifying for the third extraction, add twice the amount of hydrochloric acid (1 to 3) necessary to neutralize the alkali, and repeat the extraction with two 100-cc portions of ether. Place the neutral, alkaline, and acid extracts in adust-free atmosphere; and allow the ether to evaporate spontaneously, after which dry the residues to constant weigh over sulphuric acid, using flat-bottomed dishes 2 J inches in diameter, 1} inches in height, and of about 100 cc capacity. The dishes should be thoroughly cleaned, wiped dry, and allowed to stand in a sulphuric acid desiccator at least two hours before weighing. In order to avoid the generation of static charges of electricity, they should not be wiped immediately before weighing. Run two blank determinations with each series of ether extracts, and deduct the average gain in weight of these two blanks from the weights obtained in the other determinations. SULPIIATED ASH. Weigh accurately 0.5 to 1 gram of the color into a wide platinum dish, moisten with concentrated sulphuric acid, and ignite cautiously, avoiding spattering; moisten the residue repeatedly with sulphuric acid and ignite until all the carbon is removed and a white or reddish ash is obtained. This is finally ignited at a fairly bright red heat, cooled, and weighed. The aqueous solution of this ash should be neutral to litmus, and may bo used in a quantitative test for potassium. METHODS OF ANALYSIS FOR CERTIFIED COLORS. 215 CALCIUM. Digest the residue from the sulphated ash determination with hydrochloric acid, render the solution alkaline with ammonium hydroxid, filter, and precipitate the calcium in the filtrate with ammo- nium oxalate. The precipitate of calcium oxalate is filtered on a tared gooch, washed, dried at 100° C. and weighed as calcium oxalate, the calcium being calculated from the formula CaC 2 4 .H 2 0. This method is applicable only in cases where the amount of calcium is very small. When the amount exceeds 0.5 per cent it should be determined by digesting the sulphated ash with ammonium sulphate solution made acid with hydrochloric acid, and precipitating the iron by an excess of ammonium hydroxid. The precipitate is washed, dissolved in hydrochloric acid, and reprecipitated, the filtrate and washings being added to those obtained from the first precipitate. The calcium is precipitated in the combined filtrates with ammonium oxalate, the oxalate being filtered, washed, ignited, and weighed in the usual way; the residue is weighed as oxid after ignition to bright redness. SULPHUR. Determine upon 0.2 to 0.3 gram portions by the Carius method (Gattermann, Practical Methods of Organic Chemistry, 1901, p. 81), using 3 cc of fuming nitric acid (sp. gr. 1.5), and heating the sealed tubes to 300° C. for at least eight hours. NITROGEN. Use the method of Dumas (Gattermann, Practical Methods of Organic Chemistry, 1901, p. 85). PONCEAU 3R. MOISTURE. Dry 1 to 2 grains of the finely ground dye at 109° to 110°C. in a current of dry hydrogen to constant weight. TOTAL INSOLUBLE MATTER. Determine as under Naphthol Yellow S, page 211. \o\Yoi.ATILE OR INORGANIC INSOLUBLE MATTER. Determine as under Naphthol fellow 8, i>.iL r « i 211. SODir.M OHLOBID (SESKEB \\i> KATHEWSOX'fl Million). Mix 2 grama of dye thoroughly with from l } to 3 grama of sodium carbonate, moisten with water to form a paste, again mix. dry, and ignite at a low red heat. By moistening, drying, and reigniting a 216 COAL-TAR COLORS USED IN FOOD PRODUCTS. more complete destruction of organic matter is obtained. Break up the charred mass, introduce into a 200-cc graduated flask with about 100 to 150 cc of hot water, and add an excess of potassium perman- ganate to oxidize sulphids. Destroy the excess of permanganate by adding sulphur dioxid solution until the red color changes to brown, then cool the mixture and make up to the mark with water. Filter through a dry paper, acidify 100 cc of the nitrate with nitric acid, and precipitate the chlorin as silver chlorid by the addition of silver nitrate. If the solution should be brownish from a trace of organic matter the silver chlorid does not readily coagulate and tends to pass through the filter. In this case a few drops of potassium permanganate solution are added to this acid mixture, the organic matter being almost instantly oxidized. The mixture is then decolorized with a few drops of sulphur dioxid solution, the silver chlorid is separated on a tared gooch, washed, ignited, and weighed in the usual manner. The following method may also be used for Ponceau 3R: Dissolve 5 grams of the dye in 15C cc hot water, wash into a 250-cc graduated flask, and add 25 cc of a 10 per cent solution of barium nitrate. Cool the mixture, make up to the mark, and filter through a dry paper; acidify 100 cc of the nitrate, representing 2 grains of color, with nitric acid, and treat with silver nitrate solution, the precipi- tated silver chlorid being separated, washed, ignited, and weighed in a tared gooch crucible in the usual way. SODIUM SULPHATE. Dissolve 2 grams of dye in 100 cc of hot water, wash into a 200-cc graduated flask, add 50 grams of pure sodium chlorid, cool, and make up to the mark with water. Filter through a dry filter, dilute 100 cc of the filtrate to 300 cc, acidify with hydrochloric acid and precipi- tate the sulphates with barium chlorid. Filter, wash, ignite, and weigh in a tared platinum gooch in the usual way. HEAVY METALS. Determine as given under Naphthol Yellow S, page 212. ARSENIC. Determine as given under Naphthol Yellow S, page 212. ETHER EXTRACTIVES. Determine OS given under Naphthol Yellow S, page 214. SULPHATED ASH. Determine as given under Naphthol Yellow S, page 214. METHODS OF ANALYSIS FOR CERTIFIED COLORS. 217 CALCIUM. Determine as given under Naphthol Yellow S, page 215. SULPHUR. Determine as given under Naphthol Yellow S, page 215. NITROGEN (SEEKER AND MATHEWSOX's METHOD). Treat 2 grams of the color with 25 cc of a saturated solution of sulphur dioxid and 1 gram of zinc dust and warm the mixture gently until it becomes colorless. This should take place in from two to three minutes, but if it does not add more sulphur dioxid solution in small portions at a time until the color is destroyed. Then add 30 cc of con- centrated sulphuric acid and 0.7 gram of mercuric oxid or its equiva- lent of metallic mercury and digest the mixture, make alkaline, and distil as directed on page 6, Bulletin 107, Revised, Bureau of Chemistry, United States Department of Agriculture, under the Kjeldahl process. CRUDE CUMIDIN. Dissolve 20 grams of dye in 400 cc of hot water and pour the solu- tion, a little at a time, into a reducing solution composed of 75 grams of stannous chlorid dissolved in 180 cc of concentrated hydrochloric acid. Heat the mixture on a steam bath until it is straw colored, cool, add an excess of sodium hydroxid, and extract in a separatory runnel with ether. Separate the ether layer and distil off the solvent until the residue measures about 50 cc. Then cautiously heat over the steam with constant agitation until the odor of ether disappears, after which the last of the moisture is removed by introducing a few pieces of solid caustic soda and allowing to stand. The residue con- sists of crude cumidin and should boil above 220° C. Cumidin nitrate is sparingly soluble in water. ORANGE I. MOISTURE. Determine as given under Ponceau 3K, page 215. TOTAL INSOLUBLE MATTER Determine ;>^ given under Naphthol Yellow 8, page 211. NONVOL \ I ill. OB INORGANIC EN80LUBLI maitku. Determine ;i- given under Naphthol Yellow s. page 211, SODD m ( BLOBJB. Determine afl given under Ponceau 3R, page 215. 218 COAL-TAR COLORS USED IN FOOD PRODUCTS. SODIUM SULPHATE. Dissolve 1 gram of dye in 100 cc water contained m a 200-cc grad- uated flask and treat the solution with 60 cc of a 20 per cent solution of potassium chlorid. Make the mixture up to the mark with water, shake, filter through a diy paper, dilute an aliquot of 100 cc of the filtrate to 200 cc, acidify with 1 cc of 10 per cent hydro- chloric acid, treat with 5 cc of 10 per cent barium chlorid solution, and allow to stand over night. If a precipitate has been formed this is separated, ignited, and weighed in the usual way. HEAVY METALS. Determine as given under Xaphthol Yellow S, page 212. ARSENIC (SEEKER AND SMITH' S METHOD). Dissolve 2 grams of dye in a mixture of 130 cc water and 70 cc of 95 per cent alcohol. Add about 10 cc of strong bromin water to convert any arsenite to arsenate. Make the mixture alkaline with a few cubic centimeters of strong ammonium hydroxid and add, from a pipette, 20 cc of a sodium phosphate solution containing 100 grams of crystallized sodium phosphate per liter, after which mag- nesia mixture (containing 55 grams of hydrated magnesium chlorid, 55 grams of ammonium chlorid, and 88 cc of ammonium hydroxid, sp. gr. 0.9, per liter) is added from a burette, stirring vigorously. The amount of magnesia mixture to be added should be in slight excess of that necessary to precipitate the phosphate completely, and should be previously ascertained by a blank experiment. Finally add 10 cc of ammonium hydroxid (sp. gr. 0.96) and allow the whole to stand for at least eight hours. Separate the precipitate by filtra- tion and wash it free, or nearly so, of dye with a mixture of one-third alcohol and two-thirds water containing one-tenth its volume of ammonium hydroxid (sp. gr. 0.90) . Dissolve the precipitate from the paper with 20 per cent sulphuric acid, the washings being collected in a large porcelain crucible. Add 5 cc of concentrated nitric acid to the contents of the crucible and evaporate the whole almost to dryness. The mixture need not be colorless at this point, a brown colored solution giving equally accurate results. Add 20 cc of water to the residue in the crucible and then 10 cc of a saturated solution of sulphur dioxid. Evaporate the solution to a syrupy consistency to remove sulphur dioxid, and then take up in 20 cc of water, place in a 30-cc evolution bottle, add 5 cc of concentrated sulphuric acid, and determine the arsenic by the modified Gutzeit method as given under Naphthol Yellow 8, page 213. BTHEB EXTRACTIVES. Determine as given under Naphthol Yellow S, page 214 METHODS OF ANALYSIS FOR CERTIFIED COLORS. 219 SULPHATED ASH. Determine as given under Naphthol Yellow S, page 214. CALCIUM. Determine as given under Naphthol Yellow S, page 215. SULPHUR. Determine as given under Naphthol Yellow S, page 215. NITROGEN. Determine as given under Ponceau 3R, page 217. TEST FOR ORANGE II (SMITH AND MATHEWSON's METHOD). The following solutions are required : (1) Fifteen per cent titanium trichlorid. (2) Freshly prepared diazotized sulphanilic acid. Heat a mixture composed of 1 gram of sulphanilic acid, 10 cc of water, and 20 cc of concentrated hydrochloric acid on a steam bath for five minutes with occasional shaking. Cool to about 10° C. and add slowly 10 cc of a 1 per cent solution of sodium nitrite. Allow the mixture to stand at about 10° for 10 minutes, dilute to 1 liter, and shake until all the solid particles have dissolved. (3) Stannous chlorid, prepared by dissolving 40 grams of stannous chlorid in 100 cc of concentrated hydrochloric acid. Dilute this ten times with water, immediately before using. Place 2 cc of a 0.1 per cent solution of the dye in a colorimeter tube having a capacity of about 100 cc, add a small drop of titanium trichlo- rid solution and shake until the mixture is decolorized. Standards containing the same quantity of color composed of a mixture of pure Orange I and known amounts of Orange II are treated in the same way, the volume of the solution at this point to measure less than 5 cc. Dilute the decolorized solutions to 50 cc with 95 per cent alcohol and equalize the temperature 4 by immersing the tubes in water at room temperature. (Note. — A slight coloration that may develop at this point may bo disregarded.) Allow the tubes to stand in the water for about five minute-, add 2 cc of the diazotized sulphanilic acid, and mix thoroughly. If the titanium trichlorid has not been used in too great excess, the first few drops of the diazotized sulphanilic acid will cause a formation of color. Allow the coupling to proceed for three minutes and then add 5 CC of the diluted BtannOUS chlorid with vigorous shaking. In two minutes the blue color due t«» ( tarange I will disappear j Leaving only (he pink caused by the coupling product of Orange II. The depth of color in the tube containing the dye under examination may then he compared to the standards. 220 COAL-TAR COLORS USED EN FOOD PRODUCTS. AMARANTH. MOISTURE. Determine as given under Ponceau 3R, page 215. TOTAL INSOLUBLE MATTER. Determine as given under Naphthol Yellow S, page 211. NONVOLATILE OR INORGANIC INSOLUBLE MATTER. Determine as given under Naphthol Yellow S, page 211. SODIUM CHLORID. Determine as given under Ponceau 3R, page 215. SODIUM SULPHATE. Dissolve 2 grams of dye in 100 cc of warm water in a 200-cc grad- uated flask, and add 36 grams of pure sodium chlorid. Allow the mixture to stand with frequent shaking for one hour, and after cooling make up to the mark with a saturated salt solution. Shake the mixture and fdter through a dry paper; dilute 100 cc of the filtrate (representing 1 gram) with water, acidify with hydrochloric acid, and precipitate the sulphates with barium chlorid. The precipitate is separated, washed, and ignited upon a tared platinum gooch crucible. HEAVY METALS. Determine as given under Naphthol Yellow S, page 212. ARSENIC. Determine as given under Naphthol Yellow S, page 212. ETHER EXTRACTIVES. Determine as given under Naphthol Yellow S, page 214. SULPHATED ASH. Determine as given under Naphthol Yellow S, page 214. CALCIUM. Determine as given under Naphthol Yellow S, page 215. SULPHUR. Determine as given under Naphthol Yellow S, page 215. METHODS OF ANALYSIS FOR CERTIFIED COLORS. 221 NITROGEN. Determine as given under Ponceau 3R, page 217. LIGHT GREEN S F YELLOWISH. MOISTURE. Determine as given under Ponceau 3R, page 215. TOTAL INSOLUBLE MATTER. Determine as given under Naphthol Yellow S, page 211. NONVOLATILE OR INORGANIC INSOLUBLE MATTER. Determine as given under Xaphthol Yellow S, page 211. SODIUM CULORID. Determine as given under Ponceau 3R, page 215. SODIUM SULPHATE. Dissolve 5 grams of the color in 100 cc of water, warming by means of a gentle heat. Dissolve 1 1 grams of safranin in a separate portion of 400 cc of water, also by warming, taking care in both instances to prevent loss by evaporation. Mix the two solutions, shake thor- oughly, and niter through a diy filter. Render an aliquot portion of the filtrate alkaline with sodium hydroxid and remove the excess of safranin by shaking with two successive portions of amyl alcohol. Wash the combined amyl alcohol layers with two portions of water and add the washings to the main aqueous solution, which is then acidified with hydrochloiic acid and sulphates determined in the usual manner by preciptiation with barium chlorid as barium sulphate. JILAYY METALS. Determine as given under Xaphthol Yellow S, page 212. ARSENIC. Determine as given under Naphthol Yellow S, page 212. 1:1 mi: EXTRACTIVES. Determine as given under Naphthol Yellow S, page 21 I Si i I'll \ I I i) ASH. Determine as given under Naphthol Yellow S, page 214. 1 (II M. Determine as given under Naphthol Yellow S, page 215 222 COAlrTAR COLORS USED IN FOOD PRODUCTS. SULPHUR. Determine as given under Naphthol Yellow S, page 215. NITROGEN. Determine on 2-gram portions by Gunning's modification of the Kjeldahl process, using a little copper sulphate to assist the oxidation (see page 7, Bulletin 107, Revised, Bureau of Chemistry, United States Department of Agriculture). ERYTHROSIN. MOISTURE. Determine as given under Ponceau 3R, page 215. TOTAL INSOLUBLE MATTER. Determine as given under Naphthol Yellow S, page 211. NONVOLATILE OR INORGANIC INSOLUBLE MATTER Determine as given under Naphthol Yellow S, page 211. SODIUM CHLORID. Dissolve 5 grams of the dye in 400 cc water and acidify with dilute nitric acid. Make the mixture up to 500 cc with water, and then filter through a dry filter. Determine chlorids in an aliquot of 200 cc of the filtrate by precipitation with silver nitrate, washing, igniting, and weighing the silver chlorid in a tared gooch crucible in the usual manner. SODIUM SULPHATE. Employ another aliquot of the filtrate obtained after precipitating the color acid as above in the determination of sulphates, precipi- tating the latter as barium sulphate in the usual manner. HEAVY METALS. Determine as given under Naphthol Yellow S, page 212. ARSENIC (SEEKER AND SMITH'S METHOD). Dissolve 16 grams of dye in 370 cc of water, add 5 cc of strong bromin water, and finally 25 cc of dilute sulphuric acid (1 to 4). Shake thoroughly and filter through a dry filter. Place an aliquot of 250 cc, representing 10 grams of color from the filtrate, in a porce- lain casserole, add 5 cc concentrated nitric acid (very important to prevent loss of arsenic), and evaporate till fuming has ceased. Reduce the residue with sulphur dioxid solution, evaporate to small hulk, and determine the arsenic in the form of apparatus used by Bishop in METHODS OF ANALYSIS FOR CERTIFIED COLORS. 223 his modification of the Gutzeit test. See under Naphthol Yellow S, pages 212 and 213. (Note.— It is somewhat difficult at times to recover 250 cc of filtrate, but less may be used and a correction made, if necessary.) ETHER EXTRACTIVES. Determine as given under Naphthol Yellow S, page 214, omitting the acid extraction. SULPHATED ASH. Determine as given under Naphthol Yellow S, page 214. IODIN, BROMIN, AND CHLORIN (CORNELISON'S METHOD). Mix 0.2 to 0.3 gram of the sample with 2 grams of pure potassium bichromate and 15 cc of strong sulphuric acid in the evolution flask of an apparatus made entirely of glass, with ground-glass joints. Thoroughly mix the contents of the evolution flask, so that all lumps are disintegrated, and then heat at 100° C. for 15 minutes, after which raise the temperature to 150° C. for thirty minutes, a cur- rent of air dried over calcium chlorid and potassium hydroxid being drawn through the apparatus during this time. Iodin remains in the evolution flask as iodic acid; bromin passes off as such, and may be absorbed by allowing the air passing through the apparatus to bubble through 1 per cent sodium hydroxid. Chlorin passes out of the evolution flask as chromyl chlorid, and may also be absorbed in sodium hydroxid. Cool the mixture containing the iodic acid, and reduce the chromic acid by addition of sulphur dioxid, about 20 cc of a saturated solution being required. When enough has been added, the precipitated iodin redissolves, and the clear green color of chrome alum appears. Filter, wash the paper with distilled water, dilute the filtrate and washings to about 300 cc, and add an excess of silver nitrate. Boil till the silver iodid has flocculated, allow to stand for a few hours, and separate and weigh the silver iodid in a tared gooch. It sometimes happens that the mixture containing the iodic acid, after the reduction with sulphur dioxid, becomes turbid, owing, apparently, to separation of a basic chromium Bulphate. Very often the turbidity can not be removed by filtering, and it lias been found advisable in this case to reject the determination and begin anew. IODIN (SEEKKK AM) MATHEWSON'S METHOD). 1 Place from 0.3 to 0.4 gram of the crythrosin in a porcelain casserol, dissolve this in 5 cc of a 10 per cent Bodium hydroxid solution, then add :;."> <•<• of a 7 per cent solution of potassium permanganate. After l U. 8. . BnraM "f Chflmlstry 'reulur (>o. The tttlmattftll uf iodin in organic coiiijhhiiwLs anf water and treat the solutioo wit 1 1 lo giams <>!' pure Bodium chlorid. Aftei the salt has olved, make up the volume to exactly 200 cc. Shake the mix- • in 1 1 !<■ operation of drying particular can should be obaerved to prevent aoce a of reducing gases to the mixture. ADDENDA. 225 ture thoroughly and filter through a dry filter. Dilute 50 cc of the filtrate with 200 cc of water, acidify with 1 cc of 10 per cent hydro- chloric acid, and treat with an excess of barium chlorid solution. After standing overnight, the precipitate is separated, washed, ignited, and weighed in the usual way. HEAVY METALS. Determine as given under Xaphthol Yellow S, page 212. ARSENIC (SEEKER AND SMITH ; 8 METHOD). Treat 10 grams of dye in a Kjeldahl flask with 100 cc water and 10 cc concentrated nitric acid. Warm gently and finally boil until all action has ceased. Transfer to a beaker, make alkaline with ammonium hydroxid, and proceed from this point, as in the case of the solution of Xaphthol Yellow S, page 212, continuing from the point at which sodium phosphate is added. ETHER EXTRACTIVES. Dissolve 3 grams of dye in 200 cc of water and extract with ether. as directed under Xaphthol Yellow S. page 214. 8ULPHATED ASH. Determine as given under Xaphthol Yellow S. page 214. CALCIUM. Determine as given under Xaphthol Yellow S. page 215. SULPHUR. Determine a- given under Xaphthol Yellow S. page 215. NITROGEN. Determine as given under Light Green SF Yellowish, page 222. XVIII. ADDENDA. ADDITIONAL EXAMINATION OF COAL-TAR DYES. Since the foregoing report was written the chemical examination of the coal-tar dye specimens collected in the summer of 1907 has been completed. This investigaf ion shows that 7 of the ( rreen Table num- bers then reported were not, mi i he market . although those funushing the specimens included them at, the time that the samples were sup- plied. These numbers are as follow-: not physiologically examined, n... . etto L5 24 52 17:; 25 34 171' 44 3 190 .7. L57 12 B 74 i, 191 24, eo 171 1 1 id 228 OAL-TAR COLORS USED IX FOOD PRODUCTS. l J a^ Rose:.- 54 v. Raumer. Santori 47, 158 Schacher: 45, 75, 191 Schultz 56 Society of Swiss Analytical (hem- ifitfl 43, 75 Stilling 55, 56, 57, 180 Canton of Tessin 75 Tschirch 43, 57 Page. 37 Weyl 23, 24, 26, 4 i, 58, 181 Weyl (Leffman translation) 14, 37, 74, 179, 181 ^ft'illiams, H. W 55 Winogradow 52 Winton 24, 59 Young 1 50 O ) ss>' lliliiiii 3 1262 09216 9118