He paused and we waited expectantly.

“I have had the body exhumed and have repeated the tests. The alkaloid which I discovered had given precisely the same results as in their tests.”

My heart sank. What was he doing—convicting the man over again?

“There is one other test which I tried,” he continued, “but which I can not take time to duplicate to-night. It was testified at the trial that conine, the active principle of hemlock, is intensely poisonous. No chemical antidote is known. A fifth of a grain has serious results; a drop is fatal. An injection of a most minute quantity of real conine will kill a mouse, for instance, almost instantly. But the conine which I have isolated in the body is inert!”

It came like a bombshell to the prosecution, so bewildering was the discovery.

“Inert?” cried Kilgore and Hollins almost together. “It can’t be. You are making sport of the best chemical experts that money could obtain. Inert? Read the evidence—read the books.”

“On the contrary,” resumed Craig, ignoring the interruption, “all the reactions obtained by the experts have been duplicated by me. But, in addition, I tried this one test which they did not try. I repeat: the conine isolated in the body is inert.”

We were too perplexed to question him.

“Alkaloids,” he continued quietly, “as you know, have names that end in ‘in’ or ‘ine’—morphine, strychnine, and so on. Now there are two kinds of alkaloids which are sometimes called vegetable and animal. Moreover, there is a large class of which we are learning much which are called the ptomaines—from ptoma, a corpse. Ptomaine poisoning, as every one knows, results when we eat food that has begun to decay.

“Ptomaines are chemical compounds of an alkaloidal nature formed in protein substances during putrefaction. They are purely chemical bodies and differ from the toxins. There are also what are called leucomaines, formed in living tissues, and when not given off by the body they produce auto-intoxication.