1865.] Sorby on Detection of Blood-Stains by Spectrum- Analysis. 205 



5. Application of Polarized Light with the Spectrum. 



In employing polarized light with the spectrum, it is requisite to 

 insert a Nicol's prism between the prism and the achromatic con- 

 denser, and to use another Nicol's prism as analyzer, over the eye- 

 piece. If then, the polarizer and analyzer being so placed as to givo 

 a dark field, a plate of selenite or other doubly refracting crystal bo 

 so arranged on the stage of the microscope as to bring its maximum 

 power of depolarization into play, we see a spectrum of the usual 

 colours, crossed at equal intervals by black interference bands, shaded 

 off quickly on each side to the neighbouring colour. The number of 

 these bands depends on the thickness of the plate. Thus, if, when 

 examined with an ordinary polariscope, the colours are of the second 

 order, there is one dark band in the spectrum ; if of the sixth order, 

 three bands ; and if of higher orders, still more. But, as is well 

 known, high orders are very faint reds and greens, and still higher 

 mere white, so that a great difference in thickness makes no change in 

 colour. When, however, examined with the spectrum and polarized 

 light, the dark bands are very visible, and increase in number with 

 the thickness of the selenite, until so numerous as to be undistinguish- 

 able. If, instead of comparing different thicknesses of the same 

 mineral, we compare much the same thickness of different minerals, 

 of course making allowance for difference in the position of the prin- 

 cipal axis of the fragments, the number of the bands indicates a differ- 

 ence in the power of double refraction, and therefore this character 

 becomes useful in studying some branches of microscopic mineralogy. 

 For example, microscopic fragments of quartz, topaz, and calcite, of 

 much the same size, show respectively a very few, a moderate num- 

 ber, and very many bands ; and in studying thin sections of rocks, 

 two minerals which do not show any other characteristic difference, 

 might be distinguished by this means. 



6. Detection of Blood-Stains. 



Passing from these almost purely scientific questions, I will now 

 describe the application of the spectrum-microscope to a very practi- 

 cal subject, viz. the detection of blood-stains in criminal inquiries. 

 The optical properties of blood have been described by Hoppe,* 

 and in still greater detail by Professor Stokes.")" Hoppe suggested 

 that the peculiar and characteristic spectrum of fresh blood might be 

 employed as a test, but the plan he proposed was to moisten the blood- 

 stain with water and examine it direct ; which of course could not bo 

 done, if it were on a highly-coloured fabric. Professor Stokes also 

 suggested that the spectrum of deoxidized haBmatin might be employed 

 for the purpose, but did not further develop the subject, being evi- 

 dently desirous to investigate it chiefly in relation to chemistry, optics, 

 and physiology. However, in considering the question, I soon be- 

 came convinced that it deserved a most careful study with special 



* 'Virchow'a Archiv.' vol. xxiii. (1862) 446. 

 f ' Proceedings of R. S.' xiii. (1864) 355. 



q2 



