“Come, Dennis, don’t be standin’ there lookin’ at the roof. The roof’s not goin’ to hurt ye. Ye’re not down here to be lookin’ after the roof. I’ll be doin’ that. Just ye ‘tend to yer shovelin’.” It was the voice of Cavanaugh near at hand. Unconsciously McGlathery had stopped and was staring upward. A small piece of earth had fallen and struck him on the back. Suppose! Suppose! Know, O reader, that the business of tunneling is one of the most hazardous and dramatic, albeit interesting, of all known fields of labor. It consists, in these latter days at least, in so far as under-water tunneling is concerned, of sinking huge shafts at either end or side of a river, lake or channel (one hundred feet, perhaps, within the shore line) to a depth of, say, thirty feet below the water level, and from these two points tunneling outward under the bottom of the river until the two ends meet somewhere near the middle. The exact contact and precise joining of these outer ends is considered one of the true tests of skilful engineering. McGlathery personally understood all this but dimly. And even so it could not cheer him any. And it should be said here that the safety of the men who did the work, and the possibility of it, depended first on the introduction at either end, just at the base of the shafts and then at about every hundred or so feet, as the tunnel progressed outward, of huge cylindrical chambers, or locks, of heavy iron—air locks, no less—fifteen feet in diameter, and closed at each end by massive doors swinging inward toward the shore line, so that the amazing and powerful pressure of air constantly forced outward from the shore by huge engines could not force them open. It was only by the same delicate system which causes water locks to open and close that they could be opened at all. That is, workingmen coming down into the shaft and desiring to pass into the head of the tunnel beyond the lock, would have to first enter one of these locks, which would then gradually be filled with air compressed up to the same pressure as that maintained in the main portion of the tunnel farther in. When this pressure had been reached they could easily open the inward swinging door and pass into the tunnel proper. Here, provided that so much had been completed, they might walk, say, so much as a hundred or more feet, when they would encounter another lock. The pressure in the lock, according to who had last used it, would be either that of the section of the tunnel toward the shore, or of the section beyond, toward the centre of the river. At first, bell cords, later telephones, and then electric signals controlled this—that is, the lowering or raising of the pressure of air in the locks so that one door or the other might be opened. If the pressure in the lock was different from that in your section, and you could not open the door (which you could not), you pulled the cord or pushed the button so many