The serum drawn off at each bleeding must be severely tested, which is done by gauging its antitoxic power _in vitro_, when mixed with venom, and also its preventive effect. An antivenomous serum may be considered to be utilisable when a mixture of 1 c.c. of serum with 0·001 gramme of cobra-venom produces no intoxicating effect in the rabbit, and when a preventive subcutaneous injection of 2 c.c. of serum into a rabbit of about 2 kilogrammes enables it to resist, two hours later, subcutaneous inoculation with 1 milligramme of venom. The _preventive power_ may be very quickly tested by injecting a rabbit, _in the marginal vein of the right ear_ for example, with 2 c.c. of serum, and injecting, _five minutes afterwards_, _in the marginal vein of the left ear_, 8 milligramme of venom. This dose of 1 milligramme generally kills the control rabbits in less than thirty minutes when introduced into the veins, and in from two to three hours when injected beneath the skin. This rapid proof by _intravenous injection_ is extremely striking and demonstrative; it can be effected in public during a class or lecture in less than an hour, and enables an immediate estimate to be formed of the value of an antivenomous serum. When it is intended to adopt this method, it is essential to make use of a recent solution of venom, for solutions from a week to a fortnight old, although sterile, have already lost a large portion of their toxicity, and, if these be employed, the dose of venom calculated to kill the control animals in thirty minutes, for example, takes an hour or more to do so. I always prepare my test solutions of venom in the following manner:-- Ten milligrammes of dry cobra-venom are weighed in a delicate balance. The venom is dissolved in 10 c.c. of 0·8 per cent. physiological salt solution, which takes a few minutes. When the venom is thoroughly dissolved it is transferred to a test-tube, which is immersed for three-quarters of an hour in a water-bath heated to + 72° C. In this way the non-toxic albumins are coagulated without modifying the neurotoxic substance. The solution is poured on to a filter of sterilised paper, and the clear liquid which is collected is immediately put up in glass phials, which are hermetically sealed, or in small sterilised bottles. Its toxicity is tested upon control animals, and it may be kept for five or six days if protected from light, or for several weeks in a refrigerator at about 0° C. _One-tenth of this solution corresponds exactly to 1 milligramme of dry venom._ As for the antivenomous serum, as soon as its antitoxic value has been ascertained by the methods that I have just described, and it has been separated from clots and red corpuscles by suitable decantation, it is portioned out, with the usual aseptic precautions, into small sterilised bottles of 10 c.c. capacity, without the addition of any antiseptic.