458 Proceedings of Royal Society of Edinburgh. [sess. 
however, that protection lasts for at least a considerable period of 
time, even when the last protective dose has not been a large one. 
For example, to a rabbit which had last received twice the mini- 
mum-lethal dose of Crotalus venom, the same dose was administered 
twenty days subsequently, and it altogether failed to produce any 
toxic symptoms. 
Before passing to the next part of my communication, it may be 
stated that as yet no sufficient data have been obtained for affording 
an explanation of these remarkable facts. It is obvious that the 
blood of protected animals must contain some substance or sub- 
stances which are not present in the non-protected animals, by 
which the lethal and toxic effects of venoms are prevented. I have 
observed that when the blood-serum of protected animals is added 
to a solution of venom, a distinctly observable reaction occurs, and 
this reaction may be of significance when considered along 'v^ith 
circumstances, which will be stated in the remaining part of this 
communication, and especially with the circumstance that the 
blood-serum itself possesses but little physiological activity. This 
protective substance may be produced in the body by the influ- 
ence of the venom, but it is also conceivable that the substance is 
actually a part of the venom itself, which gradually accumulates 
under repeated administrations, whereas the lethal and toxic con- 
stituents of the venom are more rapidly destroyed or eliminated. 
Having thus succeeded in producing a high degree of protection 
in animals against the toxic effects of serpents’ venom, the blood- 
serum of these animals was, in the next place, collected for the 
purpose of testing its antidotal properties. In this portion of the 
investigation, the method followed was essentially the same as that 
described in a communication made by me to this Society in 1871, 
on “The Antagonism between the Actions of Physostigma and 
Atropia,” as it appeared to be the most direct method for obtaining 
accurate knowledge of the value of an antidote. 
A few preliminary experiments were early made with the 
serum of animals in whom the protection had not been carried 
to a high degree, and they were sufficient to show that antidotal 
properties are possessed even by this serum. It soon became 
apparent that in order to obtain some reasonable approximation to 
constancy in the conditions of the experiments, it was necessary 
