and Vegetable Colouring -matters by the Spectrum Microscope. 145 



tution, Friday, March 4th, 1864 *. It immediately occurred to me 

 that a spectroscope might be combined with a microscope, and em- 

 ployed to distinguish coloured minerals in thin sections of rocks and 

 meteorites. I was soon led to examine many other coloured sub- 

 stances, and found that the instrument is more useful in connexion 

 with qualitative analysis when only very small quantities of mate- 

 rial can be obtained. 



2. Apparatus. 



I have an ordinary large binocular microscope, and use an object- 

 glass of about three inches focal length, corrected for looking through 

 glass an inch thick, the lenses being at the top, so as to be as far as 

 possible from the slit. This is placed at the focus ; and between it 

 and the lenses, at a distance of about half an inch from them, is a 

 compound prism, composed of a rectangular prism of flint-glass, and 

 two of crown-glass of about 61°, one at each end. This arrangement 

 gives direct vision and a spectrum of the size most suitable for these 

 inquiries, since a wide dispersion often makes the absorption-bands 

 far too indistinct. In order to be able to compare two spectra side 

 by side, a small rectangular prism is fixed over half the slit, and 

 with the acute angle parallel to and just passing beyond it. This 

 gives an admirable result, the only defect being that, when the 

 spectra are in focus, their line of junction is some distance within it ; 

 and therefore to correct this I use a cylindrical lens of about two 

 feet focal length, with its axis in the line of the slit, which can easily 

 be fixed at such a distance between the slit and the prisms, as to 

 bring the spectra and their line of contact to the same focus. In 

 front of the slit, close to the small rectangular prism, is a stop with 

 a circular opening, to shut out lateral light, and a small achromatic 

 lens of about half an inch focal length, which gives a better field, 

 and counteracts the effect of the concave surface of the liquid in the 

 tubes used in the experiments, if they are not quite full. These are 

 cut from barometer-tubes, having an internal diameter of about one- 

 seventh of an inch, and an external diameter of about three-sevenths 

 of an inch. They are made half an inch long, ground flat at each 

 end, and fixed with Canada balsam on slips of glass two inches long and 

 about six-tenths of an inch wide, so that the centre of the tube is about 

 one-fourth of an inch from one edge. By this arrangement the liquid 

 may be examined through the length of the tube by laying the slip 

 of glass flat on the stage of the microscope, or through the side of 

 the tube by placing the slip vertical and the tube horizontal. Cells 

 of this size can be turned upside down and deposits removed without 

 any liquid being lost ; and the upper surface of the liquid is suffi- 

 ciently flat, even when inclined at a considerable angle. If requisite, 

 small bits of thin glass can be laid on the top, which are held on by 

 capillary attraction, or may be fastened with gold-size, if it be de- 

 sirable to keep the solution for a longer time. When the depth of 

 colour is too gi-eat in the line of the length of the cell, we can at 

 once see what would be the effect of about one-fourth of the colour 

 * Phil. Mag. S. 4. vol. xxvii. p. 388. 



Phil. Mag. S. 4. Vol. 31. No. 2.28. Aug. 1867. L 



