What I propose is to provide every seagoing submarine with one or more mobile submersible bases of supply in the form of boats without motive power of their own which can be towed by the military under-water boat and sunk upon the sea-bed at convenient points where they will best suit the purposes of the subaqueous torpedo vessel. Naturally you ask what would happen if submarine scouts should sight a submarine towing a convoy of this sort. Wouldn't the submarine have to desert her supply vessels and sink alone beneath the surface? My answer is no. Of course, this assumes that the submarine at the time is traversing waters that are not too deep for her to go to the bottom. She would take her tender or tenders down with her under such circumstances, for the supply boats would be built to stand safely the test submergence of the military submarine; that is, a depth of two hundred feet. The question may arise as to how I can control the sinking of the crewless consorts, holding as they would only supplies, and having none of the operative mechanisms that constitute a necessary functional part of the fighting undersea boat. This is illustrated by the method with which I controlled the submergence of a tender with which I salvaged the coal from a sunken barge in Long Island Sound years ago. That cargo boat had tanks into which water could be admitted from the sea, and certain of the inlet passages were closed by means of check valves which were automatic, seating themselves by the tension of springs. In order to submerge the boat it was only necessary to admit water purposely and to open a valve on the deck for the escape of the air as the water entered. To refloat the tender after it had reached the bottom and was loaded, a diver went from the submerged _Argonaut_ by way of the diving compartment and attached a hose to the deck vent. This hose was then connected to the compressed-air flasks in the submarine. The air was blown down through the pipe into the ballast tanks and the water forced outboard, past the check valves that yielded in that direction, but reseated and closed themselves as soon as the air pressure stopped. In this fashion buoyancy was reacquired and the tender rose to the surface. Of course, the initial sinking operation required the presence of someone in a small boat alongside the tender which I have just described. This would not be feasible in the case of the military supply boats I have in mind. These must be made to sink by suitable controlling devices manipulated from within the military craft, but in principle the cycle would not differ from that which I have outlined.