slightly longer, which gave me the key to the problem. I admit that I should not have known what these things were, unless I had seen them under different conditions a short time previously.
When riding to the hospital one morning I noticed on the horizon a long, narrow dark cloud. By about 12 o'clock the cloud had reached the town, and a dense mass of locusts was passing over, and millions were settling on the ground, on the trees, and on the roofs of the houses. And when in about thirty-six hours the visitation passed on, every blade of grass and every green leaf, with the exception of a few noxious weeds, had disappeared, and the ground was as brown and bare as one of the streets of London. But as the locusts ate, they had necessarily to defecate, and their stools consisted of spindle-shaped bodies, which were composed of the fibrous, indigestible parts of the grass matted together, and also of the silicious spicules which are to be found in many grasses. I had noticed them lying about after the locusts had gone and had examined them, and it was quite clear to me that those which had been deposited on the large expanse of barrack roof had become partly disintegrated, and had been washed into the tanks with the first shower of rain. The specific gravity of these particles kept them in suspension for some time, so that they sank slowly, and it was evident that I had to deal with a case of irritant mechanical poisoning caused by the passage of these spicules through the alimentary canal, similar to, but not so severe as, the ground glass poisoning of the Obeah Man in the West Indies (or the diamond dust of the Egyptians).
That this view was correct was shown by the sequel. I stopped the source of water supply, and utilised a surface well in the barrack-yard, the water from which was carefully boiled. The roof and the gullies were cleaned, and a considerable quantity of similar debris was found. The
