He pushed the bottle across to Thorndyke, who carried it to the window, and, extracting a small quantity of the contents with a glass rod, examined the pasty mass with the aid of a lens; then, lifting the bell-glass cover from the microscope, which stood on its table by the window, he smeared a small quantity of the suspected matter on to a glass slip, and placed it on the stage of the instrument.

“I observe a number of crystalline particles in this,” he said, after a brief inspection, “which have the appearance of arsenious acid.”

“Ah!” ejaculated Mr. Barton, “just what I feared.  But are you certain?”

“No,” replied Thorndyke; “but the matter is easily tested.”

He pressed the button of the bell that communicated with the laboratory, a summons that brought the laboratory assistant from his lair with characteristic promptitude.

“Will you please prepare a Marsh’s apparatus, Polton,” said Thorndyke.

“I have a couple ready, sir,” replied Polton.

“Then pour the acid into one and bring it to me, with a tile.”

As his familiar vanished silently, Thorndyke turned to Mr. Barton.

“Supposing we find arsenic in this arrowroot, as we probably shall, what do you want us to do?”

“I want you to come and see my brother,” replied our client.

“Why not take a note from me to his doctor?”

“No, no; I want you to come—­I should like you both to come—­and put a stop at once to this dreadful business.  Consider!  It’s a matter of life and death.  You won’t refuse!  I beg you not to refuse me your help in these terrible circumstances.”

“Well,” said Thorndyke, as his assistant reappeared, “let us first see what the test has to tell us.”

Polton advanced to the table, on which he deposited a small flask, the contents of which were in a state of brisk effervescence, a bottle labelled “calcium hypochlorite,” and a white porcelain tile.  The flask was fitted with a safety-funnel and a glass tube drawn out to a fine jet, to which Polton cautiously applied a lighted match.  Instantly there sprang from the jet a tiny, pale violet flame.  Thorndyke now took the tile, and held it in the flame for a few seconds, when the appearance of the surface remained unchanged save for a small circle of condensed moisture.  His next proceeding was to thin the arrowroot with distilled water until it was quite fluid, and then pour a small quantity into the funnel.  It ran slowly down the tube into the flask, with the bubbling contents of which it became speedily mixed.  Almost immediately a change began to appear in the character of the flame, which from a pale violet turned gradually to a sickly blue, while above it hung a faint cloud of white smoke.  Once more Thorndyke held the tile above the jet, but this time, no sooner had the pallid flame touched the cold surface of the porcelain, than there appeared on the latter a glistening black stain.