J. A. Rossman, of the firm of Rossman & Bracken, of New York, writes:

"After reading Mr. David S. Cowan's letter, and seeing his diagram for the kitchen and laundry boiler system, in the RECORD, I conclude that his whole trouble is caused by the position of the stopcock under the kitchen sink. He says the object is to use the laundry range during the summer, and appears to understand the necessity of shutting off the source of that supply when the range is not in use. He must certainly do the same with the other range and boiler when they are not in use. You will readily see by the position of the stop-cock, which he must use, that it would prevent the street supply from coming to his kitchen sink, as well as to his kitchen boiler. If he should place a check valve where he proposes, he would still have to close the stop-cock under the sink to prevent the cold water from passing through the kitchen boiler into hot-water system. He then would get no cold water to the kitchen sink or his bathroom.

"If he will place a stop-cock where he proposes to put the check valve - viz , on the cold-water branch to kitchen boiler, where it should have been put originally, he will be able to overcome his trouble. I think if he placed a check valve where he proposes, and in the autumn when he abandons the use of the laundry range and shuts that source of supply off he will very soon be asking you why his boiler or range pipes burst, or some other weaker point in his hot-water system gives way, because there would be no possible chance for expansion.

"I assume from the drawing, of course, that this is a direct-pressure system, and there is no tank in the house where a relief pipe could be taken. It is possible that my conclusions are wrong, from the fact that your drawing may be misleading, as it would hardly seem credible that any plumber would put two stop-cocks on supply a with no branches between, where in using either of them you would necessarily shut the entire system off the house, excepting the laundry tubs as it was originally fitted, and not have one on cold supply to boiler. According to your drawing, in order to shut the water off the kitchen boiler, you are obliged to shut it from the kitchen sink as well as the bathroom and whatever fixtures there maybe upstairs. Referring to your comments thereon, I can hardly see how a check valve placed as he proposes would interfere with the circulation of the hot-water system, if there was one in the building properly arranged. I mean by that, taken from a high point of the main hot-water riser and returned to the bottom of the boiler. The check would have no effect upon the circulation whatever. I also fail to see why you should favor your proposed 'second method' and abandon the supply pipe b, which you consider 'more ample and vigorous ' than a, which you retain, except on the theory that two wrongs make one right/'