14 On some Improvements in the 



dissolves out liaematin, and therefore, in testing portions of soil, 

 they should be digested in considerably more of that solvent than 

 will fill an experiment cell, and after the solution has become quite 

 clear it should be concentrated by evaporation. The spectrum of 

 deoxidized haematin may then be seen by following the ordinary 

 method. The same process should be adopted in examining stains 

 on clothes impregnated with earth or earthy dust, and marks on 

 iron contaminated with much rust, if water will not dissolve out 

 unaltered blood or methaemoglobin. 



The importance of being able to detect blood-stains on leather 

 was prominently brought before me by a case in which the trial of 

 a suspected person depended on the nature of certain dark marks 

 on his gaiters. The presence of tannic acid so completely mordants 

 the blood, that neither water nor citric acid will dissolve it, and am- 

 monia gives rise to a most inconveniently dark solution. If the 

 stain is on the surface, and has never been wetted, a thin shaving 

 should be cut off, so as to have as much blood and as little leather 

 as possible, and the blood should be dissolved off without exposing 

 the solution to the action of the leather itself. This may be accom- 

 plished by taking one of the experiment cells, nearly filled with 

 water, bending the shaving, and inserting it into the upper part of 

 the tube, so as to touch the water, being careful to arrange it so that 

 the stain may be on the convex side of the leather, and in contact 

 with the water. When a drop of blood falls on leather, many red 

 globules are filtered out from the serum and left on the surface, 

 and, when thus treated, they dissolve, and the coloured solution 

 sinks at once to the bottom of the cell, without coming in contact 

 with the leather. The various spectra may then be observed in the 

 usual manner. This method would be of little or no use if the 

 stain had been wetted, and for a long time I concluded that after 

 such treatment it would be impossible to recognize blood. How- 

 ever, after many experiments, and after having again and again 

 almost given up the inquiry in despau', I found that the diflficulty 

 could be overcome in a very simple manner. The best solvent for 

 the insoluble compound of the colouring matter of the blood with 

 tannic acid, is hydrochloric acid diluted with about fifty times its 

 bulk of water. If stronger or weaker, the result is not so good. 

 When a portion of unstained common brown leather is digested in 

 this dilute acid, the solution is scarcely tinged yellow. On adding 

 excess of ammonia, the colour becomes pale purple, or neutral tint, 

 made deeper when the double tartrate and the ferrous salt are 

 added, but remaining nearly clear. This gives a spectrum very 

 dull all over, but without any trace of definite bands in any part. 

 The depth of colom' varies much with difierent specimens of 

 leather. A portion of similar material soaked with wetted blood, 

 gives a yellow solution, made brown-purple and turbid by the 



