I went to work, built a house that in several important features was quite contrary to the prevailing ideas of what a poultry house should be, and used it for nearly a year before saying anything publicly about it.

This house was a mere shell or shed; the walls were of common hemlock boards, laid perpendicularly on a light frame, and the joints between the boards covered on the back and ends of the house with common battens; the joints on the front were left open. The roof was of shingles, laid on strips of furring placed three inches apart.

The house was not tight anywhere. As I used it the first winter, it having been built in a hurry late in the fall,— the battens were merely held in place with two or three small nails in each, and were loose enough to let a great deal of air in around them. The cracks in the boards, some quite large, were not covered at all. The front of the house had double doors six feet wide in each section, and these were kept open all day unless a storm would beat in, and all night except on very coldest nights or nights when storms would drive into the doors.

The house was built on wet ground, - that is, ground that was thoroughly soaked by the late rains. After the roof was on, the ground in the house was spaded up; and when the house, a few days later, was ready to put the fowls in, the surface of the ground in it was dry, but a mere scratching of it would show damp earth. By spring there was about two inches of dry earth on top, and the soil damp below that. The walls in the house were dry, never a bit of moisture on them except as a driving rain might wet through the joints and cracks. This would dry out quickly, and I never noticed any ill effects from it.

The house was cold, the temperature in it was but little higher at any time than that out doors; yet going into it I noticed that it always felt comfortable, with a feeling like what you get in a warm, sheltered spot out doors,—not at all like the warmth of a heated building.

The hens always seemed as comfortable in it as I ever saw hens anywhere. There were some fifty to sixty in it that winter, and only two slight cases of colds, which recovered

immediately with no treatment but a single application of vaseline to the head. The egg yield was fair, comparing favorably with average good reports. The hens in this house were Light Brahmas. I published a report of experi

ence with this house, and on the strength of my experience with it began to urge correspondents who had much trouble with damp houses or unhealthy fowls to open up their houses, and keep them open at all seasons.

The next year I continued to use this house the same way, but put in a part of it a pen of Silver Dorkings, a breed reputed to be rather delicate and susceptible to cold, and having combs which would be more quickly affected by the frost. I also built another smaller house, on the same principle, and put in it a brood of late-hatched Brahma chickens. The Dorking male had his comb very badly frozen, but the hens' combs were scarcely nipped at all. They were put into the house December 1, began to lay in about three weeks, and laid well all winter.

The brood of chickens hatched June 27 that I put in the small house in October made a remarkable record for growth and early maturity. I got the first egg January 15, and by February 1 all the pullets, nine in number, were laying;

and from that time until February 22, when I took some out to put in my breeding pen, the nine pullets were giving me about a fifty per cent egg yield, and still growing. I weighed them all on February 22, and found five of the nine

DETAILS OF CONSTRUCTION OF HOUSE SHOWN ON P. 398. This house is 8 feet square on the ground, 4 feet high at sides, 7 feet in the middle. Cost about $12. A, sill plan, with position of corner boards indicated at cccc, dddd; E, construction of a corner; B, side; C, front; D, method of cutting pattern for rafters.

weighing nine and one-half pounds or over, and none of the others much below it. These pullets were not fat; they were big pullets, in good condition. Their house, as shown in one of the illustrations accompanying this article, was battened only on the back and half way forward on each side. The door was open practically all the time, day and night; and the windows were always partly open except when a storm came against one, when that one would be closed.

The next winter, still continuing to use the houses I had already built, I built on frozen ground late in November a house very similar to that in which the Brahma pullets just mentioned had been kept. It was made of old material, and was a little poorer in construction all around than the other house, joints between sides and roof not as good, and some very wide joints between boards in front. Into this house I put my Dorking hens and pullets. This winter, 1903-04,

was, as all know, a record-breaker for cold and snow. Frequently after a night of driving storm I would go to this house and find so much snow sifted through the joints at the back that there was a light sprinkle of it all over the litter on the floor, so much that the hens would not come down from the roost until the litter had been shaken up; and the backs of the hens as they sat on the roosts were well coated with snow. The tips of the hens' combs were somewhat frosted (no male was put in the house), but these hens gave me about a forty per cent egg yield in January, 1904. There were very few hens in any kind of house anywhere doing better at that time.

During the several winters covered by the experiments mentioned my fowls were cared for by myself when at home, at other times by different members of the household, none of whom had any particular interest in or skill in feeding fowls. Our one rule for feeding was to be sure that the hens had an abundance. In the first half of the winter I had to be away so much that I found it impossible to keep accurate egg records. When I relied on others, they forgot; and so I gave up thought of making statistical figures complete, concluding that circumstances limited me to general demonstrations of a few leading facts, and the exact results possible in cold houses and comparison of these with results in other houses would have to be left to others. My trials demonstrated that hens could be kept healthy and giving average good egg yields in cold, open houses. So far as I could judge, they consumed no more food than when kept in warm houses, though theoretically it should have required more.

Last winter the only item of experience in my poultry yard having a further bearing on this subject was the performance of a pen of July-hatched Single-combed White Orpingtons. These were put in an old poultry house that was sheathed, papered and covered with common lapped siding on the sides and with shingles on the roof. The front of this house was about half glass, an immovable window, and there was a half-window in each end, the door being at the north-west corner. The growth of an apple tree near the house had forced a board from the front next the roof,

and there was an opening here the length of the house wide enough "to throw a cat through." The west window was wide open all winter. The pullets in this house had been sold early in January, and, as I was expecting to ship them any day, we kept no record of their laying. The buyer failed to take them; but, as I still intended to let them go, and as some weeks had passed with no records kept, we let them go along unrecorded. Through January, February and March these pullets, nine in number, laid rarely less than seven eggs a day, and often nine for several days in succession. Nearly all their combs were somewhat frosted, and the comb of the male at one time quite badly frosted.

Now, of what use was it to demonstrate that warm houses are not essential to egg production, and that hens can be kept healthy and productive in very cold houses?

I went to the extreme, giving my fowls houses that were mere shelters, to show more convincingly, by extreme illustrations, that warm houses, which are more expensive to construct and require more careful attention to operate, were not absolutely essential. My tests, though not furnishing statistics, do show conclusively that egg production is not necessarily dependent upon "spring" conditions; and that the cold, open house for poultry is the style of house in which the labor of caring for them can be reduced to the minimum. Considered in connection with the general difficulties with tight buildings, they indicate also that the safest and most profitable and practical type of house for most poultry keepers is the house that is so constructed that it does not require close attention from the poultry keeper to keep conditions in it safe. They have not developed what is the best construction of house. It is reasonable to assume that a little better construction than I have used would be better, would afford more protection, without making conditions that interfere with the steady renewal, in abundance, of supplies of fresh air.

Between the house so tightly built that the ventilation in it is very bad, and one so open that the temperature in it is but slightly higher than out doors in extremest cold weather, there is a medium form of construction and an intermediate