In so far as the coloring matter was examined, the following table shows the character of the pigments used and the relative number of times they respectively were found:

In connection with the coloring matter, however, it should be remembered that in the great majority of cases no attempt was made to distinguish them further than to determine whether they were of an organic or inorganic nature. Only one analyst, Weber, determined the nature of the coloring matter in each instance. Two of the number, Stubbs and Wallace, did not report the number of samples colored.

On examining the above list it will be noticed that in 56 cases the name of the coloring matter is given. Furthermore we learn that in 2 cases a coal tar color was detected, and in 3 instances some fluorescent color, but the names are not given. The work of analyzing these 250 samples was divided among 10 chemists, each of whom made twenty-five analyses. As these chemists were widely separated from one another and chose their samples from their home markets, the result of their analyses show the composition of confections from various parts of the United States. Regions as widely separated as California, Louisiana, Nebraska and Massachusetts are represented in this report.

The general summary of the work of these chemists is sufficient to show that the coal tar colors are very extensively employed for coloring confections. All but four of the eighteen coloring matters given in the above list are properly classed as coal tar colors.

My own work although confined to the examination of 21 samples shows that the coal tar colors are used just as freely at the present time as they were five years ago. The following table shows the results of the analyses made by my assistant Mr. C. S. Brinton and myself.

Sixteen of the 21 samples of candy examined were colored with coal tar colors. Sample No. 6, a green candy made to represent pods of green peas, contained a blue coal tar color, indigo carmine, and a yellow coloring material, the character of which I was unable to determine. It will be noticed that Rhodamine B was found in seven of the twenty-one samples examined by us, while this pigment is not reported at all as occurring in the candies examined by the chemical division of the U. S. Department of Agriculture. On the other hand, Bengal red, a somewhat similar coal tar pigment occurs five times in the report of the Department of Agriculture, but does not appear in our list. It may be that Rhodamine B is supplanting the use of Bengal red, particularly in this section of the country.

The report of Bulletin No. 13 shows about 70 per cent. of confections to be colored with coal tar dyes. Of the 21 samples examined by us, 80 per cent contained coal tar colors.

In addition to the sucrose which was present in all the candies examined, nearly all contained glucose and many also starch and dex'rin. No evidence was found of any excess of mineral matter or any injurious substance, unless perhaps the coal tar colors should be so considered. The 21 samples examined by us were obtained in Philadelphia, Reading and West Chester.

In this connection a few words of explanation in regard to the manufacture of the coal tar colors may be of interest.

In Wagner's Hand Book of Chemical Technology, the coal tar colors are classified as follows: 1. Analine colors; 2. Carbolic acid colors; 3. Naphthaline colors, and 4. Anthracene colors. While this classification does not include all coal tar colors, it has the advantage of simplicity and is sufficient for popular purposes. Among the many substances obtained from coal tar are benzene, the starting point of the analine colors; carbolic acid, the starting point of the carbolic acid colors; naphthaline, from which the naphthaline colors are obtained, and anthracene, the basis of the anthracene colors.

One of the most important of the analine colors is fuchsin, known also as magenta, and aniline red. The various steps in the preparation of this dye are briefly described as follows: By the distillation of coal tar is obtained a small amount of a liquid called benzene. By the action of nitric acid on benzene is obtained a liquid called nitrobenzene. If iron filings and acetic acid be added to nitrobenzene, and the mixture subjected to distillation, analine oil will be produced. From this last named substance the analine colors are manufactured. Fuchsin can be made by heating analine oil with nitric acid. Crude fuchsin has a greenish black color and appears to the eye somewhat like bituminous coal. It is often seen in pieces of considerable size.

Some study has been devoted by different investigators to the physiological action of the coal tar colors for the purpose of determining

whether or not their use for coloring foods is objectionable. In most countries coloring of food with substances injurious to health is forbidden by law. In some cases the laws relating to adulteration of food, name the colors that are allowable and those that are forbidden. Among the coloring substances forbidden by the laws of France, is quite a long list of coal tar colors, many of which are now used quite extensively in this country for coloring foods. As example of this may be cited, magenta, methyl orange, eosin, fast yellow.

The Austrian government positively prohibits the use of all coal tar colors for coloring substances included in the list of foods.

The term food in its legal sense includes confections and condiments of every kind.

Professor Weyl has made experiments on dogs with 31 coal tar colors. As a result of these experiments he concludes that 6 of this list of 31 are poisonous and 3 suspicious; the remaining 22 are nonpoisonous, at least to dogs. The 6 poisonous colors are naphthol green B, picric acid, saffron substitute, Martiu's yellow, orange II and metanil yellow. The colors named as suspicious are aurantia, soudan I and chrysoidin. Several of the pigments passed by Prof. Weyl as non-poisonous are capable of producing very unpleasant effects, as will be seen from the following statements taken from Weyl's work on the Sanitary Relation of the Coal Tar Colors; translated by Dr. Henry Leffmann:

"Of the 23 azo colors subjected to examination only 2, metanil yellow and orange II, produce such effects when administered by the stomach that we can consider them poisonous. With dogs the lethal dose is less than one gram per kilo of body weight of orange II, and only 53 gram per kilo of metanil yellow. Of the remaining colors some produce vomiting (e. g., Bismark brown), and others diarrhae (fast brown, chrysamin R), and many develop slight albuminuria."

Weyl's entire list contains only one coloring substance (viz: Bismark brown), which I have found in candy. Martiu's yellow (a poisonous pigment), however, has been found by me in two samples of mustard and soudan I (a suspicious pigment), in butter color.

Prof. H. A. Weber, of Ohio, has made a series of experiments with four coal tar colors for the purpose of learning their influence upon digestion. The following is a brief summary of his results.

Concluding his report, Prof. Weber says:

"It seems then so far as these four colors are concerned that none interfere with both peptic and pancreatic digestion, but that each color interferes seriously with either the one or the other. What the action of other coal tar colors may be, can, of course not be inferred from this limited number of experiments, but it may be safely said that bodies which have such a decided action in retarding the most important functions of the animal economy can not properly have a place in our daily food and drink."

Experiments in artificial digestion were made by Mr. Brinton and nyself with seven of the coal tar colors. The material to be digested was powdered beef, previously dried and the fat extracted with petrolic ether. The fluid for peptic digestion consisted of 100 c.c. of a 2-10 of 1 per cent. solution of hydrochloric acid, containing .020 grams of pepsin. The table shows the results of these experiments.

The effect of three of these colors on pancreatic digestion is shown in the following table:

The pancreatic fluid used in these experiments was made as follows: Water, 100 c.c.; sodic bicarbonate pancreatin.

It will be observed that in every case the presence of a coal tar color exerted some retarding influence on digestion. The results indicate that eosin and acid magenta interfere decidedly with peptic digestion, and that methyl violet produces just as decided interference with pancreatic digestion.