of the methods of experiment adopted by him in the investigation of the Solar Spectrum and of the Spectra of the Elementary bodies, and an explanation of the results obtained, together with a theory respecting the constitution of the Sun's body.

The subject is divided under five heads, viz. :

4. The Chemical Constitution of the Solar Atmosphere.

5. The Physical Constitution of the Sun.

The original Memoir was printed in the Transactions of the Physical Section of the Berlin Academy for 1861, and is accompanied by two lithographic maps of the fixed lines in a portion of the Solar Spectrum, showing the coincidence of these with the bright lines of certain metallic Spectra. Impressions of the same maps, which are exact copies of Professor Kirchhoff's drawings, and masterpieces of lithographic art, have been secured for the English edition, and this will, therefore, resemble in every detail the German text.

The maps of the remaining portion of the Spectrum between the lines A and G, not represented in the two Plates accompanying the Memoir, will be published as soon as Professor Kirchhoff has completed the observations and the drawings.

OWENS COLLEGE, MANCHESTER,

February, 1862.

H. E. R.

ON THE SOLAR SPECTRUM.

Ir a pure solar spectrum, obtained by a prism, be observed by means of a telescope of small magnifying power, an indistinct maze of fine lines and nebulous bands is seen between the more prominent lines which Fraunhofer has designated by separate letters. If several prisms and a higher magnifying power be employed, and if the apparatus be constructed with the requisite care, a number of groups of lines are seen which are so characteristic and can be so easily recognised, that they may, with propriety, be compared to the groups of stars by means of which the various constellations are distinguished. In Fraunhofer's original map of the solar spectrum' but very few of these groups can be recognised; and the same remark applies to the drawing on a much larger scale more recently published by Brewster and Gladstone. I have endeavoured to produce a drawing of these groups of lines in the brightest portion of the solar spectrum, which shall represent the reality as faithfully and completely as possible. Plates I. and II. are lithographic copies of my own drawings3.

The magnificent apparatus, which I have employed for the observations on the spectrum, was manufactured in the celebrated optical atelier of C. A. Steinheil of Munich. It is represented in Plate III. fig. 1.

Philosophical Transactions for 1860, p. 149.

My drawing is intended to include that portion of the spectrum contained between the lines A and G. I must, however, confine myself at present to the publication of a part only of this, as the remainder requires a revision, which I am unfortunately unable to undertake, owing to my eyes being weakened by the continual observations which the subject rendered necessary.

The telescope A is screwed on to a circular iron plate, having its surface turned plane; the eye-piece of this telescope is replaced by a metal plate containing an accurately made slit, formed of two knife-edges. The breadth of this slit can be regulated by a micrometer screw, and the slit itself can be brought into the focus of the achromatic object-glass by means of a rack-and-pinion motion. The focal length of this lens is 18 paris inches, and its diameter 18 paris lines. The telescope B, containing a similar object-glass, is fastened to a brass arm, which is moveable about the centre of the iron plate, and this motion can be given to it either with the hand or with a micrometer screw. Between the two object-glasses four flintglass prisms are placed; each of these possesses circular refracting faces 18 paris lines in diameter; three of them have a refracting angle of 45°, the fourth one of 60°. Each of the prisms is cemented on to a small brass stand furnished with three set-screws. The telescope B has a double motion with respect to its brass arm; it can be moved about a horizontal axis, and can be drawn in and out in the direction of this axis, as is seen in the figure. The magnifying power of the telescope which I have used in my observations is about 40.

I have employed the following method for setting up the apparatus. The telescope B adjusted for viewing a very distant object, then placed in a line with the tube A, and the slit pushed in or out until it was plainly seen through the object-glass of A. The two tubes were next brought into such a position that their axes cut one another at right angles, and in the point of intersection was placed a plate of glass with plane and parallel surfaces, fastened on to a brass stand like the prisms. The setscrews of this stand were so altered, and the telescope B inclined in such a direction, that the image of the middle point of the slit always coincided with the point of intersection of the cross wires of the telescope B, whichever side of the plane plate of glass was turned towards the observer. The glass plate whose surfaces were thus placed at right angles to the surface of the iron plate, was then exchanged for one of

the prisms, and the set-screws of its stand were altered until the same coincidence appeared, when first one and then the other of its refracting surfaces served as the reflector. The refracting edge of the one prism being thus placed at right angles to the surface of the iron plate, the positions of the other prisms were similarly arranged. A heliostat, furnished with clockwork motion, was then placed so that the sun's rays fell into the observing room, and the iron plate was moved until the beam which passed through the slit illuminated the central portion of the object-glass of the tube A. The prisms and the telescope B were then arranged so that the spectrum was plainly seen; and the prisms were turned into the position of minimum deviation for the rays which had to be observed, and so placed that the rays which passed through the centre of the object-glass of the tube A also passed through the middle of the prisms. The telescope B was, lastly, moved in a direction parallel to itself, until those of the rays under examination which passed through the middle of the object-glass of A, likewise fell upon the middle of the object-glass of B. This condition, which was found to be essential to the purity of the spectrum, was secured by means of two caps, each of which contained a narrow vertical slit, and could be placed before each object-glass. When the telescope B was wrongly placed, the whole field appeared dark; by gradually moving the tube a band of light was seen which could be brought into the centre of the field. When the caps were taken off the objectglasses, the slit made of the right degree of breadth, and the eye-piece of B slightly altered, the spectrum appeared with a degree of distinctness which I believe had never before been attained.

In order to measure the distances between the several lines, I made use of a circular divided scale, attached to the head of the micrometer screw, by means of which the telescope B is moved. The eye-piece was placed so that the cross wires made angles of 45° with the dark lines; the point of intersection was brought by means of the micrometer screw to coincide with each of these lines, the divisions read off, and the estimation of the breadth and darkness of the line noted.

After these observations the drawings of the lines were prepared. For the purpose of drawing the lines I employed a kind of dividing instrument, consisting of the scale of a cathetometer placed in a horizontal position, on to the slide of which I had a tracer fitted such as is used in ordinary dividing machines. In the place of the graver a common drawing-pen was fastened, and to the screw of this a circle was attached, marked so as to correspond to the breadth of the lines to be drawn. These lines were also arranged according to their degree of blackness. First the darkest lines were drawn with thick black indian-ink; the ink was then diluted to a certain extent, and the lines of the next shade drawn, and so on to the lightest series. As soon as a portion of the spectrum had been drawn in this manner, it was compared with the actual spectrum, the mistakes in breadth and darkness of the lines, as well as in their position, corrected by fresh estimations, and the drawing made anew. A second comparison and another drawing were then made, and this process repeated until all the groups of lines appeared to be truthfully represented.

Above the drawing of the spectrum I have placed a scale divided into millimetres, and having an arbitrary commencement. This is, in the first place, useful for the purpose of obtaining an easy means of nomenclature for the lines. I propose that each of the lines should be called by the number placed above it on the scale. Thus, for instance, I designate the two lines, which Fraunhofer called D, by the numbers 1002.8 and 1006.8; Fraunhofer's line E consists of the two lines 1522'7 and 15237; the three lines 1633°4, 1648′3, and 1655'0 are in Fraunhofer's map called b. By means of this scale we are likewise enabled

'The Plates I. and II. are printed with ink of six different tints from six different stones. Owing to mechanical difficulties it was found to be impossible to effect a complete coincidence between the separate copies and the original. In order to enable the reader to judge of the slight errors introduced by the lithography and the printing, I give, in a Supplement, the series of readings of the scale which in the original drawing correspond to the middle of the various lines. For the sake of easy comparison I have added to this table the values of darkness and of breadth of the lines. The degrees of darkness