( 9 ) 



IV. — On some Imjjrovefnenis in tJie Spectrum Metliod of Defecting! 

 Blood. By H. C. Soeby, F.E.S., &c. 



In the following paper I sliall give a condensed account of what I 

 have been able to learn in connection with this subject, and omit 

 everything that does not bear directly on determining whether any 

 stain is, or is not, due to hlood. There does not appear to be any 

 probability of our being able to decide by this means whether it is, 

 or is not, human. 



The spectrum-microscope used in these inquiries should have a 

 compound prism, with enough, but not too great, dispersive power, 

 or else the bands would be as it were diluted, and made less dis- 

 tinct. A combination of two rectangular prisms of crown glass, 

 with a rectangular of very dense flint, and another of less dense, of 

 such an angle as to give direct vision, turned towards the sht, as 

 lately made for me by Mr. Browning, appears to be the proper 

 medium, and has other important advantages. The cells used for 

 the experiments should be made from barometer tubing, and be 

 about one-eighth of an inch in internal diameter, and half an inch 

 long, one end being fastened to a piece of plate glass with purified 

 gutta-percha, like an ordinary cell for mounting objects in liquids. 

 It is, however, a very great advantage to insert between the plate 

 and the cell a diaphragm of platinum foil, having a cu-cular hole 

 about two-thirds of the internal diameter of the tube, fixed so that 

 its centre corresponds with that of the cell. This prevents any 

 light passing upwards that has not penetrated through the whole 

 length of the solution, which is very important when using direct 

 concentrated sunlight to penetrate through turbid or very opaque 

 liquids. A small spatula made of stout platinum wire, flattened at 

 the end, is very convenient for adding small quantities of the re- 

 agents ; and they should be stirred up in the cells with a platinum 

 wu'e, flattened and turned up square at the end, like a small hoe. 

 The reagents commonly employed are a somewhat diluted solution 

 of ammonia, citric acid, the double tartrate of potash and soda, used 

 to prevent the precipitation of oxide of iron, and the double sul- 

 phate of the protoxide of iron and ammonia, employed to deoxidize ; 

 but in some special cases diluted hydrochloric acid, carefully-purified 

 boric acid, and sulphite of soda are required. 



The character of a stain varies much with its age, and with the 

 nature of the substance on which it occurs. If quite recent, and if 

 the substance has no immediate influence on blood, the stain would 

 contain little or no colouring matter but haemoglobin. This is 

 easily dissolved by water, and when properly diluted — neither too 

 strong, nor too weak — it gives the well-known spectrum, with two 

 dark absorption-bands in the green. The addition of a very little 



