When engaged, during the course of last summer, in the examination of the action of various medicinal and poisonous agents on blood, I observed the remarkable change in the colour of the blood of animals poisoned with nitrite of amyl. In every case in which I induced death by this agent the blood presented a dirty chocolate coloration, instead of the characteristic colour of normal blood. When examined by means of the spectroscope, this blood was found to present optical characters very different from those of ordinary blood, for the two well-marked absorption bands of 0 – hæmoglobin, or (as the readers of Professor Stokes' paper will know them better as) of scarlet cruorine, were found to be extremely faint, and only visible when comparatively thick layers of fluid were examined; whilst an additional and very faint absorption band in the red, occupying the same position as that of acid hæmatine, was seen.