ZOOLOGY AND BOTANY, MICROSCOPY, ETC. 493 



which can be clamped together on the same axis at any desired angle. 

 The first mirror throws the beam of light, issuing from the projection- 

 lens, direct on to the first slit of the spectral apparatus. The light 

 emergent from the second slit is then reflected by a fixed mirror on to 

 the second adjustable mirror, and thence into the observation tube. 



Spectroscope with Adjustable Dispersion.* — P. Kriiss found that 

 the desirability of this instrument was suggested by the requirements of 

 the dyeing industry. Most colouring materials have characteristic spectra, 

 and it is, therefore, obviously necessary to make the spectrum as clear 

 and as perfect as possible. It is found, however, that the spectrum is 

 dependent on several factors, e.g. the most suitable solvent, the proper 

 degree of concentration of that solvent, and the thickness of the layer 

 placed before the slit of the instrument. But, in addition to these fairly 

 obvious factors, the spectrum of a dye usually requires a suitable dis- 

 persion. A dispersion suitable for one medium may be so unsuitable for 

 another that the absorption bands may be indistinct and quite un- 

 recognisable. At present the operator keeps several spectroscopes of 

 various dispersive powers ready, but this inconvenience the author seeks 



Fig. 58. 



to remedy by the design of his iustrument. He places two direct-vision 

 prisms behind one another, and rotates them equally in opposite directions 

 about their common optical axis. This arrangement is similar to that 

 adopted by Abbe for his refractometer. The two opposed rims R, R^ 

 (fig. 58) of the tubes carrying the prisms P and P l5 are provided with 

 engaging toothed wheels, and between these wheels is a driving wheel T, 

 by which the two tubes are rotated through exactly equal angles in 

 opposite directions. The two prisms P and P l5 are perfectly similar to 

 one another, and are direct-vision for a ray of medium wave-length. If 

 one prism has a dispersion D, then the total dispersion will be 2 D. If 

 each is rotated through 90°, then the total is zero. Thus the limits He 

 between 2 D and 0. If <f> be the angle of rotation, measured from the 

 position in which the refracting edges of both prisms are perpendicular 

 to the direction of the slit, and if d be the length of the spectrum 

 measured, from the ray which passes without deviation to a ray of 

 desired wave-length, then the combined dispersion parallel to the slit is 

 zero, and perpendicular to the slit 2 d sin <f> ; hence only the latter has 

 to be measured. If the absorption-band attained in a single-prism 



* Zeit. f. Instr., xxvi. (1906) pp. 139-42 (2 figs.). 



