R. Haddow, Winnipeg, Man., writes: "With reference to the correspondence on fresh-air inlets in The Engineering RECORD, the experience of another plumber in Winnipeg might not be out of place. Bearing in mind that the thermometer goes down to 40 degrees and 50 degrees below zero, I never have found a single instance wherein a fresh-air inlet froze over with hoar frost with properly adjusted drain and soil pipes. I think Mr. Hughes must be making a mistake, because it is not the fact of cold air rushing in a pipe that freezes up the mouth of the pipe; it is the warm vapor coming out of the soil pipe at top that freezes up But even that I believe can be obviated, by enlarging the pipe at its terminus, and have as little pipe as possible after going through roof, leaving mouth of pipe clear and open.

"But the greater trouble of all with us is to get a fresh-air inlet that will not freeze up the main trap, and that I think you cannot have where you enter your pipe immediately over the trap. Mr. McAr-thur's idea of the cold air striking the bend before entering the trap I fail to see, because the cold air never has time to strike on that point, from the fact that the suction of the soil pipe is so great. If it was forced in, then it would be another thing. I think it it is more from the fact that he put the inlet further from the trap than it was before that kept it from freezing, so that I think by having as much fresh-air inlet pipe inside of house (as is workable according to circumstances) as possible, and entering one or even two lengths of soil pipe from trap, will keep the trap from freezing. I use generally 2-inch pipe and carrv it 7 or 8 feet up the wall; this is to prevent it getting covered over with snow. These few remarks are drawn from practical experience."

[Letters based on experience are always welcome. Our correspondent seems to indorse the views frequently expressed as to location of fresh-air inlets in cold climates.]