powers, is excellent. It merely consists (see fig. 15) of a conical tube, turned out of boxwood, and fastened on to the microscope tube by means of a screw, h. Combined with this funnel tube is a wooden tray, in which the frame, e, d, f, g, easily slides backwards and forwards. Thin panes of glass are let in the cassette into this frame. Whilst the image is being adjusted, the thin glass, e, d, stands over the tube, and the prepared plate is put under the little cover at g, f. If the picture is well defined, the frame, e, d, g, f, is pushed into the tray, so that the part, g, ƒ, can stand over the microscope tube, and by a simple arrangement the photographic plate can be exposed. Direct sunshine will, in most cases, be necessary, and the rays should be transmitted through a cell containing the ammonia-sulphate of copper. If, however, it is desired to photograph with high powers, the plan recommended and employed by Dr. Woodward, of the Army Medical Museum, Washington, is perhaps the best. The camera box and table are both dispensed with, and the operating room itself is converted into a camera. A room is selected having a southern aspect; the window is provided with shutters on the inside to exclude light, sufficient being admitted through one or two yellow panes to enable the operator to move about freely; a small yellow pane is also let into one of the window-shutters to enable the operator to watch the face of the sky. The microscope is placed horizontally, and a heliostat outside the window throws the direct rays of the sun on to the mirror. The frame of the plate-holder runs on an iron track, ten feet long, and laid on the floor at right angles to the plane of the window. There are most ingenious arrangements for working, although at a distance, the fine adjustment of the microscope; the sun's rays pass through a solution of the ammonia-sulphate of copper.* The fixing of the picture upon the plates, the method of printing from the negatives, &c., are all extremely simple operations (especially to those accustomed to chemical manipulations), and are well described in standard treatises on photography. § 58. Colour. It will often be necessary to ascertain the exact colouring-matter used to make articles of food attractive, more especially confectionery, jellies, pickles, &c. The question will generally resolve itself into deciding as to whether the colour is harmless or poisonous, and, if the latter, whether the poison is in sufficient quantity to injure the consumer's health. The poisonous colouring-matters are those containing lead, copper, arsenic, chromium, and zinc, all of mineral origin; together with a few * The whole arrangement is figured and described in Dr. Beale's work on the microscope. injurious organic colouring substances, such as gamboge and picric acid. The non-poisonous colouring-matters are some of the aniline colours, so long as they are pure, and contain no arsenic-saffron, turmeric, annatto, chlorophyll, and generally (with some exceptions) all organic colours obtained from the vegetable and animal kingdoms. The first thing for the analyst to ascertain is whether the colouring material is insoluble or soluble in water, for, as a rule, with the exception of gamboge, the harmless colours are soluble, while the mineral are insoluble in water. The organic colours are also bleached by a solution of hypochlorite of soda. The aniline colours are soluble in alcohol. The search for poisonous matters more properly belongs to, and is treated of, in the author's work on "Poisons." With the exception of salts of lead and copper in small quantities, they are rarely met with in food, and even in the matter of confectionery, of late years, there has been a great improvement. As a rule, sweetmeats in England are not coloured with injurious matters.* The analyst having settled that the colouring-matter is one of organic origin, by its being bleached by sodic-hypochlorite, and by its solubility in water or alcohol, will next proceed to study its spectroscopic characters, either by using a pocket spectroscope, or the micro-spectroscope already described. Mr. Sorby makes use for his instrument of little cells, cut from barometer tubing. They are half-an-inch long, and with an external diameter of somewhat under half-an-inch; they are ground flat at each end, and cemented with Canada balsam near one edge of a glass plate, so that they may be examined sideways or endways. In examining an unknown colouring-matter, he adopts the following divisions: 1. Soluble in water, and not precipitated by alcohol. 2. Soluble in water, but precipitated by alcohol. 3. Insoluble in water, but soluble in alcohol. This is the more necessary to state clearly, since, on the Continent, very erroneous ideas prevail. Thus, in the Dictionnaire des Alterations et Falsifications, par M. A. Chevallier et M. Fr. Baudrimont, Paris, 1878, the adulterations of half a century ago are enumerated; and the reader is informed that the English confectioners not only falsify their sweetmeats with plaster, lime, starch, baryta, but frequently employ bronze powder, the leaf foil of copper, tin, and carbonate and arsenite of copper, verdigris, chromate of lead, red lead, and vermilion; and, further, that nearly all the ginger lozenges contain lead. Similarly, in Dr. Hermann Klencke's Lexicon der Verfälschungen, in the article, "Conditorwaaren," it is stated that almost all the English confectionery contains lead salts, often to the extent of one and a-half per cent.!! All this is nonsense. Such adulterations have been found, it is true; but instead of being common, they are rare and exceptional. 4. Insoluble in water and alcohol. He next subdivides his main divisions according to the action of bisulphite of soda. The organic colouring-matters most likely to be found may be treated of in the order of the spectrum, beginning with the red. § 59. REDS.-The common reds are-cochineal, aniline reds, alkanet, and the madder-colours alizarine and purpurine. Cochineal.-Cochineal is a red complex colouring - matter, secreted by certain species of a peculiar family of insects feeding on the Cactus coccinellifera, C. opuntia, C. tuna, C. pereskia. The chief colouring-matter of cochineal is "carminic acid," the formula of which appears to be C17H18010 By the action of dilute acids carminic acid splits up into sugar, and a beautiful colour known as carmine-red, thus Carminic acid. Water. Carmine-red. Glucose. Cochineal imparts its colouring-matters both to alcohol and water, and is precipitated by acetate of lead, carminate of lead being one of the constituent parts of the precipitate. The solutions of cochineal are purplish-red to crimson, turning a more or less rich violet-purple with alkalies, and becoming of a yellow colour on the addition of acids. The colour is well-known to chemists, as it is much used as an indicator for acids, being especially useful in titrating an alkaline liquid containing carbonates, since carminic acid is not affected by carbon dioxide like so many other colouring-matters. Cochineal in neutral solutions gives absorption-bands, but not very definite when examined by the spectroscope; if, however, it be made ammoniacal, then there are bands which differ in position only slightly from the absorption-bands of blood. No. 18 (fig. 16) is a graphical illustration of the spectrum of cochineal in water; No. 19, in alcohol; and No. 20, on the addition of nitric acid (a.) or NH, (b.). If alum is added to cochineal it loses its power of turning yellow with acids, and the purpurine band becomes so broad that the two bands almost run into each other. On addition of acetic acid they are separated, and appear as tolerably sharply-defined bands between D and E, and there is another at D. On dissolving cochineal with alum solution, a lake is obtained; on dissolving this in tartaric acid, or dilute nitric acid, the solution gives a band at b and E, and another close on D. The nitric acid solution gives a spectrum very similar to blood. An aqueous solution of cochineal may be distinguished from the red solutions of brazil-wood, sapan-wood, peach-wood, and a few others, by the fact that the calcium salt of their colouring-matters is violet, and readily soluble in water, while the calcium salt of cochineal-red is dark-purple or almost black, and insoluble in water. The following are the absorption factors for carminic acid, as obtained by using a solution containing 0.0001 grm. per cc. with a drop of NH2 : * Aniline Reds.-The aniline reds are numerous; the chief are fuchsine, safranine, and coralline. These three may be roughly distinguished from each other by adding a dilute mineral acid: fuchsine becomes yellow, safranine violet-blue, and coralline gives a yellow precipitate. Fuchsine, or Rosaniline, also called magenta, aniline red, and other names. It is a mixture of hydrochloride or acetate of para-rosaniline (triamido-triphenyl-carbinol) and rosaniline (triamido-diphenyl-tolyl-carbinol). It is distinguished from coralline, which gives a very similar spectrum (see No. 28) by the yellow colour with acids already mentioned. The absorption spectrum of fuchsine has been studied by many observers, among others by Vierordt† and by Hartley.‡ In weak solutions (0.024 mgrm. per cc.) it shows an absorption band in the visible spectrum when viewed through a stratum 4 mm. thick, extending from 25467 to 25350; in stronger solution, 0·12 mgrm. per cc., and viewed in a layer * Ueber den Einfluss der Temp. gefärbter Losungen u. die Absorption Spectren derselben zur quantitativen Spektralanalyse, von G. Krüss u. H. Krüss. Zeit. f. anorgan. Chemie, i. + Op. cit. The molecular structures of carbon compounds and their absorption spectra. Journ. Chem. Soc., li., 1857. 4 mm. in thickness, the band in the visible spectrum occupies the region from 2462 to 2580; there are also two other bands in the ultra violet-viz., one from 2300 to 2283, and another from 2247 to 2231, both, of course, invisible save with a Soret's ocular or other similar arrangement. A solution of rosaniline hydrochloride in alcohol (0.155 mgrm. per cc.) in a layer of 5 mm., gives an absorption in the visible spectrum from 2591 to 2456, and in the ultra violet from 2310 to λ274. The absorption factor for the spectrum, from a to H, has been worked out by Vierordt, and, reducing his notation to wave lengths, the following are the results from B to F: |