Thus direct injection of liquid ammonia into the veins has been found successful in the treatment of snake bite in Australia, by Professor Halford and others. Claude Bernard has shown the importance of particularly attending to the way in which the poison destroys life. For example, curare paralyzes the motor nerves, puts a stop to all motion, suspends respiration, and so brings on suffocation; yet by keeping up artificial respiration for a sufficient length of time, life may be preserved till the poison is eliminated and the danger over. Strychnia attacks the sensitive portion of the nervous system; but if the external excitement, which perpetually provokes reflex action and thereby brings on fatal convulsions, be guarded against, recovery may ensue. A frog, poisoned by strychnia, rapidly dies if frequently excited; whereas left perfectly quiet under a bell-glass, it will, cæteris paribus, recover. Lastly, we must endeavor to promote the elimination of the poison from the body, by exciting the excreting functions; for which purpose, in poisoning by arsenic, after the stomach has been well emptied, Orfila has proposed the employment of diuretics, because it has been found that this poison, like most others, is carried off in large quantities by the urine. CHAPTER IV. DETECTION OF POISONS. The detection of a poison is, in many instances, no easy matter; it should not therefore be rashly undertaken, except by one well skilled in the minutiæ of the processes to be adopted; but on the other hand, there are so many points of practical importance which may be noted by the practitioner in charge, that his attention should be specially directed to these. The exact determination of the cause of death will depend partly on the symptoms noted during life and on the appearances found after death. These come within the province of the ordinary practitioner; on the other hand, the special physical, chemical and physiological portion of the inquiry should be referred to the expert. In a book of this scope it is impossible to give full details with regard to these last, but a brief sketch nay be useful.