Professor Eccles suspected him of having engineered a sham robbery to evade a bequest to the nation. But this seemed to me a wild and unfair suspicion, and for my own part I strongly suspected the seals. Accordingly, I examined a whole series of them, minutely and exhaustively, with the result that they proved to be impressions, not of the matrix in Mr. Hollis’s ring, but of an electrotype matrix made from a wax impression.
“This new fact brought the inquiry to the next stage. It proved that the boxes had been opened and that they had been opened in Woodstock’s office. For when they came there they were sealed with Hollis’s seal, but when they left the office they were sealed with the forged seal. Things began to look rather black as regards Osmond; but, although I was retained ostensibly to work up a case against him, I kept an open mind and proceeded with the investigation as if he did not exist.
“The second stage, then, started with the establishment of these facts: a robbery had really occurred; it had occurred in Woodstock’s office; and, since the boxes had been kept in the strong-room, it was from thence that they had been abstracted. The next question was, By whom had the robbery been committed? Now, since the property had been taken from the strong-room, and since the strong-room had not been broken into, it followed that the thief must have had, or obtained, access to it. Now, there were three persons who had easy access to it: Woodstock who possessed the key, Hepburn and Osmond, both of whom occasionally had the key in their custody: There might be others, but if so, they were at present unknown to me. But of the three who were known, one, Osmond, had apparently absconded as soon as the robbery was discovered and connected with the office. Moreover, the commencement of the robberies apparently coincided in time with the date on which he joined the staff.
“Evidently, then, everything that was known pointed to Osmond as the delinquent. But there was no positive case against him, and I decided to proceed as if nothing at all were known and seek for fresh data. And my first proceeding was to make an exhaustive examination of the boxes, the wrapping-paper, and the inside packing. As to the paper, I may say that I developed up a large number of fingerprints—on the outside surface only—I never examined, as the occasion did not arise. The investigation really concerned itself with the dust from the insides of the boxes and from the packing material. Of this I collected every particle that I could extract and put it aside in pill-boxes numbered in accordance with the boxes from which it was obtained. When I came to examine systematically the contents of the pill-boxes, I made two very curious discoveries.
“First, every pill-box—representing, you will remember, one of the gem-boxes—contained one or more hairs; usually one only and never more than three. They were all alike. Each was a hair from a moustache of a light-brown colour and cut quite short, and there could be no doubt that they were all from the same individual. Consequently they could not be chance hairs which had blown in accidentally. The gem-boxes had been packed at various times, and hence the uniformity of the hairs connected them definitely with the person who packed the boxes. In short, it seemed at first sight practically certain that they were the hairs of the actual robber; in which case we could say that the robber was a man with a short light-brown moustache.
“But when I came to reflect on the facts observed I was struck by their singularity. Moustache hairs are shed very freely, but they do not drop out at regular intervals. One, two, or more hairs in any one box would not have been surprising. A man who was in the habit of pulling at or stroking his moustache might dislodge two or three at once. The surprising thing was the regularity with which these hairs occurred; one, and usually one only, in each box, and no complete box in which there was none. It was totally opposed to the laws of probability.
“The point was highly significant. Anyone can recognize a hair. Most men can recognize a moustache hair. A detective certainly could. If these boxes had been opened by the police, as Hollis had originally intended, these hairs would almost certainly have been seen and eagerly fastened on as giving what would amount to a description of the thief. They would have been put in evidence at the trial and would have been perfectly convincing to the jury.
“The more I reflected on the matter the more did I suspect those hairs. If one assumed that they had been planted deliberately, say by a clean-shaved or dark-haired criminal, their regular occurrence in every box would be quite understandable. It would be a necessary precaution against their being overlooked. Otherwise it was unaccountable. Still, the fact of their presence had to be noted and the individual from whom they came identified, if possible.
“The second discovery that I made was, perhaps, even more odd. In every one of the boxes I found particles of the fine dust which falls out of the holes in worm-eaten wood; sometimes only a few grains, sometimes quite a large number of grains, and in the aggregate a really considerable quantity.”
“But how astonishing,” exclaimed Betty, “that you should be able to tell at once that these tiny grains came from worm-eaten wood.”
“I make it my business,” he replied, “to be able to recognize the microscopical appearances of the different forms of dust. But your remark indicates a very significant point. I imagine that there will be very few persons in the world who could identify these particles in a collection of miscellaneous dust. And therein lay the value of this discovery; for if the significance of the hairs was open to doubt, that of the wood certainly was not. There was no question of its having been purposely planted. It had certainly found its way into the boxes accidentally, and the person who had unconsciously introduced it was pretty certainly unaware of its presence. It was undoubtedly a genuine clue.
“The discovery of this characteristic dust raised several questions. In the first place, how came it into the boxes? Dust from worm-eaten furniture falls on the floor and remains there. It is too coarse and heavy to float in the air like the finer kinds of dust. In a room in which there is worm-eaten furniture, you will find the particles of dust all over the floor; but you will not find any on the tables or chair-rails or mantelpiece. But these boxes must have stood on a table or bench when they were being packed and when the dust got into them. Then the dust must have been on the table or bench. But how could it have got there? It was possible that the bench, itself, might have been worm-eaten. But that was not a probable explanation, for the dust tends to fall, not to rise. It would have fallen, for the most part, from the under surface on to the floor. The most likely explanation emerged from a consideration of the next question; which was, how could one account for the large quantity that was found?
“The quantity was extraordinarily large. From the whole set of boxes we collected something approaching a quarter of a thimbleful; which seems an enormous amount if you consider that it must all have got into the boxes during the short time that they were open for packing. What could be the explanation?
“There were two factors which had to be considered: the nature of the wood and the nature of the object which had been fashioned from it; and both were important for purposes of identification. Let us consider the first factor—material. Now, these wood-boring insects do not bore through wood as the bookworm bores through paper, to get at something else. They actually feed upon the wood. Naturally, then, they tend to select the kind of wood which contains the most nourishment and which, incidentally, is usually the softest. But of all woods those of the fruit trees are richest in gum and sap and are most subject to the attacks of the worm. Walnut, pear, apple, plum, and cherry all have this drawback, and of these cherry is so inveterately ‘wormy’ that it has usually been shunned by the cabinet-maker. Now, the quantity of the wood-dust pointed to some excessively worm-eaten object and suggested one of the fruit woods as the probable material, and the balance of probability was in favour of cherry; and this was supported by the pinkish colour of the dust. But, of course, this inference was purely hypothetical. It represented the general probabilities and nothing more.
“And now we come to the second factor. What was the nature of this wooden object? A piece of ordinary furniture we could dismiss for two reasons: first, the dust from such a piece will ordinarily fall upon the floor, from whence it could hardly have got into the boxes; and, second,
