In one part of the room was a centrifugal whirler, holding a couple of test-tubes. These were filled with the fluids to be examined which contained solids in suspension, and when these tubes were whirled round at a prodigious rate the solid contents were forced to the bottom of the tubes, and could thus be readily separated. In another part of the room were test-tubes filled with serums, jellies, and meat broths of various kinds, any of which could be inoculated by touching the surface with a sterilized platinum wire which had been previously dipped in the fluid supposed to be infected by microbes. When the microbes were thus placed in their food, the test-tubes containing them would be labelled and placed in the incubator to allow the germs to multiply to their heart's content.

"Once more open the door," said Paul, smiling at his friend's amazement, and the two passed down some steps into a courtyard. All round the walls were hutches filled with guineapigs and rabbits, others contained whole families of rats and mice, some white, and some brown. Other hutches again contained cats and small dogs, while a large cage in the corner was filled with Rhesus and Bonnet monkeys. Lastly in the opposite corner was an aquarium containing a varied assortment of frogs and toads.

"What on earth do you want this menagerie for?" said Pierre.

"Why, this is the most important part of our laboratory. I will show you later what use we make of these animals. Meanwhile let us return to the first room, and we will have a chat."

"Do you always succeed in detecting the poison?" asked Duval.

"In the case of acids, alkalies, and metals or their salts, practically always, as not only are the tests easy to apply and well known, but the doses to be fatal are usually so large that one can find sufficient traces in the stomach, intestines, and liver to make a reliable test. To take an example. Here is a bottle containing what is left of the contents of the stomach of a woman who was poisoned a week ago. We have already made our report, so I can quite well use a little of what is left.

"Watch me closely. I first stir the contents well, and then filter some of it through this filter paper into this little beaker. Now I add a few drops of acid, and then allow some of the sulphuretted hydrogen gas to bubble through. Observe a bright canary yellow precipitate is forming. This shows me that arsenic is probably present. But to make quite sure I apply some further tests." Paul then poured another small quantity of the suspected fluid into a tiny porcelain dish, to which he added a few drops of pure hydrochloric acid and gently warmed it.

"Now," said Paul, "I take this slip of pure polished copper-foil and just dip it into the liquid—so, and see, it is slowly becoming covered with an iron-grey metallic film. In order to be quite sure that the coating is not due to accidental impurity, I repeat the experiment with the contents of another stomach which I know is free from any poison, and observe when I dip the foil in there is no deposit. This shows me that both the acid and the copper-foil are pure, and that in the former case the grey deposit was due to arsenic. In order to make doubly sure, I take the coated slip of copper, wash it well in water, then in ether alcohol, and gently heat it in this reduction tube. Now, let us put it under the microscope and tell me what you see."

"I see a number of shiny square crystals like little diamonds."