index, it undergoes two bendings, one at each surface of the prism; the whole ray is however not equally bent, but it is split into its constituent colours, which may be allowed to fall on a screen; the red rays will be found at one end of a continuous band of colour, the violet rays at the other; orange, yellow, green, blue and indigo being the This band of colours which are intermediate in the order named. colours is termed a spectrum. The rainbow is an instance of a spectrum produced in nature by the sun's rays passing at a certain angle A spectroscope is an instrument provided through drops of water. with a prism or prisms to enable us to obtain a spectrum artificially. In the spectrum the red rays are the least, the violet rays the most, refrangible. If the spectrum produced by one prism be immediately passed through a second prism like the first, but inverted, the coloured rays are reunited, and build up a white ray emerging from the second prism.

The different colours are due to vibrations of æther of different rates of rapidity; the wave-length of red light is greater than that of yellow, that of yellow than that of green, and so on, violet having the shortest wave-length. The wave-length of the ray changes in different media, and thus the velocity of propagation varies; the violet is the most, the red the least, retarded; the violet is thus bent the most, the red the least; hence, arises the dispersion which results in the formation of the spectrum.

In addition to the visible rays, other rays at either end are present These which can be detected by their effects, though not by the eye. rays are called respectively the ultra-red and the ultra-violet. The ultra-red rays are those in which heating effects preponderate, the ultra-violet are rays of chemical activity, i.e. produce such chemical changes as those on which the art of photography depends; the visible rays have heating, and chemical effects also, but in a subordinate degree.

The spectrum of sunlight is interrupted by numerous dark lines crossing it vertically, called Fraunhofer's lines. They are perfectly constant in position, and serve as landmarks in the spectrum: the more prominent are lettered; A, B, and C are in the red, D in the yellow, E and F in the green, G in the indigo, and H in the violet. The (in the red) and b lines (in the green) are also well marked. These lines are due to the presence of certain substances volatilised in the solar atmosphere. If the light from burning sodium or its compounds be examined spectroscopically, it will be found to give a bright yellow line, or rather two bright yellow lines, very close together; it is in fact a true monochromatic light. Potassium gives two bright

red lines, and one violet line, and the other elements when incandescent give characteristic lines, but none so simple as sodium. If now the flame of a lamp be examined, it will be found to give a continuous spectrum, like the solar spectrum in the arrangement of its colours, but unlike it in the absence of dark lines; but if the light from the lamp be made to pass through sodium vapour (produced by burning salt in an ordinary spirit flame) before it reaches the spectroscope, the bright yellow light will be found absent, and in its place a dark line, or rather two dark lines close together, occupying the same position as the two bright lines of the sodium spectrum. The sodium vapour thus absorbs the same rays as those which it itself produces at a higher temperature. Thus the D line as we term it in the solar spectrum is due to the presence of sodium vapour in the solar atmosphere. The other dark lines are also all due to the absence of certain rays, absorbed by the presence of such substances as hydrogen, calcium, barium, iron, &c., in a volatile condition in the sun's atmosphere; it being a general rule that the vapour of any material will absorb and retain light, the period of vibration of which is identical with that which it itself emits when in a state of incandescence.

The Spectroscope consists of a tube A called the collimator, with a slit at the end S, and a convex lens at the end L; the latter makes the

P

A

T

H

FIG. 26.

rays of light passing through the slit from the source of light parallel: they fall on the prism, are refracted, and then the spectrum so formed is focussed by the telescope T. The dispersion of the colours, and so the length of the spectrum, may be increased by using a train of prisms in place of P, the second prism being so placed as to receive the rays refracted from the first, the third those from the second, and so on. There are in addition various accessories to the instrument: e.g. most spectroscopes have a third tube which carries a small transparent scale

of wave-lengths; this is illuminated and is focussed by the telescope. Generally also a small rectangular prism is placed in front of the lower part of the slit at S; the solar light is focussed on to this, and we thus have two spectra, one of the candle flame or of the substance under examination below, and the solar spectrum above which can be compared with it, and lastly an image of the scale by which the position of any line or band can be read off in wave-lengths.

If we now interpose between the source of light and the slit S a piece of coloured glass H, or a solution of a coloured substance contained in a vessel with parallel sides (the hæmatoscope of Hermann, fig. 27, F), the spectrum will be found to be no longer continuous, but inter

FIG. 27. Spectroscope. A. collimator with adjustable slit at one (left) end and collimating lens at the other (right) end; B, telescope moving on graduated are divided into degrees; C, prism or combination of prisms; D, tube for scale; E, mirror for illuminating scale; F, vessel with parallel glass sides for holding fluid, shown with the flat side towards the reader; I, long spectroscope bottle for examining a deep layer of fluid; H, Argand burner; G, condenser for concen(From a photograph taken by Dr. MacMunn, from trating the light fro:n H on the slit. McKendrick's Physiology.')

rupted by a number of dark bands corresponding to the light absorbed by the coloured medium. Thus oxyhemoglobin gives two perfectly characteristic bands between the D and E lines, hæmoglobin giving only one; and other red solutions, though to the naked eye similar to oxyhemoglobin, will give characteristic bands in other positions. Chlorophyll again gives four well marked bands, especially one in the red. The study of the absorption spectra of animal and vegetable pigments is full of interest, and has been followed with most valuable results (see further chlorophyll, hæmoglobin, bile, urine, &c.).

A convenient form of small spectroscope is the direct vision spectroscope, in which by an arrangement of alternating prisms of crown and flint glass, placed as in fig. 28, the spectrum is observed by the

eye in the same line as the tube furnished with the slit; indeed slit and prisms are both contained in one tube.

The Microspectroscope is a spectroscope fitted into the ocular end of a microscope, instead of the eye-piece. There are slight variations in the instruments constructed by Browning, Hilger, and Zeiss. The

light which passes up through the microscope tube, passes through a direct vision spectroscope; the spectrum so formed is compared with a spectrum from solar light which enters by a slit in the side of the instrument, and is then by a small rectangular prism sent up also through the tube of the microspectroscope; and lastly an illuminated scale is focussed with these, as in the ordinary spectroscope.

The microspectroscope is of value in examining spectroscopically small quantities of solutions; small cells for containing the fluid to be examined are made from short pieces of barometer tubing cemented to microscope slides. In examining aqueous extracts of blood stains on garments, very often only a small volume of liquid can be obtained; in order to identify the blood pigment spectroscopically, one must here have recourse to the microspectroscope.

Dr. Mac

The instrument is also useful in examining coloured microscopic crystals, or coloured portions of microscopic organisms. Munn has made much use of the microspectroscope in this direction. He adopts the following method: a binocular microscope is taken; the microspectroscope is put in the place of one eye-piece. By means of the other eye-piece, the portion of tissue or crystal can be accurately focussed; its absorption spectrum is then seen on looking down the spectroscope in the other tube.

The absorption bands which form the characteristic features of blood and other animal liquids do not admit of having their limits determined with the same precision as is possible in the case of Fraunhofer's lines; their position in wave-lengths is usually determined in millionths of a millimetre, instead of ten-millionths. The edges of absorption bands are moreover sometimes so ill-defined, and vary so

E

much with the concentration of the solution, that often their centre is given, instead of the position of their edges.

Printed blank maps (similar to fig. 29) accompany some of Zeiss's instruments, and correspond exactly to the scale of the spectroscope. It is thus easy to draw a diagram of any given spectrum. The observer G

BC

D

E

F

commences by causing the D or sodium line to coincide exactly with that part of the scale which expresses its wave-length, that is to say, with the division 589 of the scale (this expresses the fact that the wave-length usually denoted by the Greek letter A is 589 millionths of a millimetre). Having done this, the scale is set accurately for all other points.

The usual method of determining wave-lengths, namely by interpolation curves, is thus described by MacMunn :—--1

A piece of paper ruled into square inches and tenths, obtainable from Letts & Co., has a scale of wave-lengths ruled off along the right-hand edge, and the upper edge at right angles to this has a scale corresponding to the scale of the instrument marked on it. The value of the Fraunhofer lines on the scale of the spectroscope is observed, and by a reference to Angström's numbers, their value in wave-lengths; they are then marked in their proper places on the scale with dots. A curve is then drawn through these marks as uniformly as possible. When a band or bright line has to be mapped out, all that is necessary is to take its reading on the scale; then knowing between what lines it is placed, we find its position on the curve opposite which its wave-length is printed on the right-hand edge.

THE SPECTROPHOTOMETER

The spectrophotometric method for estimating the concentration of coloured solutions was originally proposed by Bunsen and Roscoe, in 1857. In 1873 Vierordt invented a spectrophotometer, but it is Hüfner' who definitely introduced the instrument into physiological methods.

1 The Spectroscope in Medicine, p. 32.

2 Angström's calculations of the wave-lengths of the principal Fraunhofer lines are as follows in millionths of a millimetre: A, 7604; a, 718·5; B, 686·7; C, 656′2; D, 589-2; E, 5269; b, 517·2; F, 486′0; G, 4307; H1, 396-8; H2, 393′3. (Recherches sur le spectre solaire. Upsala, 1865.) Poggendorff's Annalen, vol. ci. p. 235.

4 Vierordt, Die Anwendung des Spectralapparates zur Photometrie der Absorptionsspectrum und zur quant. chem. Analyse, Tübingen 1873, and a later pamphlet in 1876.

Hüfner, Journ. f. prakt. Chemie, xvi. (1877). Zeit. physiol. Chem. vol. i. ii. vi. &c.