Fort Craig was a walled fort, made so in early days as a protection against Indians. It was typical of most of the posts at which I served in being built in the form of a square. The parade ground being a square plot varying in size at different posts, around which are located the buildings. The officers occupying one side of the square; the barracks being directly opposite and the commissary and quarter master department generally occupying one side and the commanding officer's quarters and post headquarters and adjutant's office occupying the other side. At Fort Craig just outside of these buildings was an adobe wall about ten feet high. Next to the guardhouse was an opening large enough for wagons to enter the parade ground with heavy gates to close at night, and there were some small openings in the wall for other purposes, one being near the hospital. The walls of the buildings were of adobe with heavy timbers across to support the roof of dirt. The floors were what the Mexicans called "Jaspa" (pronounced Haspa), a kind of cement made of gypsum or lime sulphate which is found in great beds through a great portion of New Mexico. It is quarried or blasted out, heated to drive out the water or crystalization, then ground into a powder and when mixed with sand and water makes a pretty fair quality of cement. It was used altogether in the floors for the military posts along the Rio Grande. The water supply at Fort Craig was obtained from the Rio Grande river and there were times about June when the snows melted in the mountains that it answered very well to a description I once read of the Missouri river water, "Too thick to drink and too thin to cultivate." This was a great bother to us during the summer rise for it was persistent for more than a month. I conceived the idea of making a filter by making a good sized ball of jaspa and charcoal which I held together by mixing a little cotton batting carefully in the mortar and kneading it into a very stiff paste. After it hardened I bored a hole in the ball and inserted a rubber tube and then put the ball in a "Tanaja," a large ungalvanized earthen jar holding eight or ten gallons of the muddy water. This jar was put in an army blanket and was swung in the hallway. The jar being porous would let enough water through to keep the blanket damp, which cooled the water. By swinging another tanaja just below the first and having it blanketed in the same way, and having a rubber tube connecting the two, I had a filter that furnished clear, sparkling, cool water. I put one in the hospital and they became quite the vogue at the post. The wood supply was brought from the mountains some thirty miles away. Trains comprising several wagons would be sent out in charge of a wagonmaster with men enough to load them promptly and by starting early and returning late they sometimes made the round trip in two days, but generally they were three days out.